🐢 Now, I have two classes. This one is a copy of the ONNX Patcher that’s used to convert ColQwen2 to ONNX format. But this one is valid only for the underlying model, that is Qwen2VLForConditionalGeneration, and ColQwen2 also has some modifications to call it.

🐇 What are these modifications?

🐢 The forward method looks something like this:

ColQwen2

    def forward(self, *args, **kwargs) -> torch.Tensor:
        kwargs.pop("output_hidden_states", None)

        # Handle the custom "pixel_values" input obtained with `ColQwen2Processor` through unpadding
        if "pixel_values" in kwargs:
            offsets = kwargs["image_grid_thw"][:, 1] * kwargs["image_grid_thw"][:, 2]  # (batch_size,)
            kwargs["pixel_values"] = torch.cat(
                [pixel_sequence[:offset] for pixel_sequence, offset in zip(kwargs["pixel_values"], offsets)],
                dim=0,
            )

        position_ids, rope_deltas = self.get_rope_index(
            input_ids=kwargs["input_ids"],
            image_grid_thw=kwargs.get("image_grid_thw", None),
            video_grid_thw=None,
            attention_mask=kwargs.get("attention_mask", None),
        )
        last_hidden_states = self.inner_forward(
            *args, **kwargs, position_ids=position_ids, use_cache=False, output_hidden_states=True
        )  # (batch_size, sequence_length, hidden_size)

        proj = self.custom_text_proj(last_hidden_states)  # (batch_size, sequence_length, dim)

        # L2 normalization
        proj = proj / proj.norm(dim=-1, keepdim=True)  # (batch_size, sequence_length, dim)
        proj = proj * kwargs["attention_mask"].unsqueeze(-1)  # (batch_size, sequence_length, dim)

        if "pixel_values" in kwargs and self.mask_non_image_embeddings:
            # Pools only the image embeddings
            image_mask = (kwargs["input_ids"] == self.config.image_token_id).unsqueeze(-1)
            proj = proj * image_mask
        return proj

🐇 How about the inner_forward method? Does it just directly call Qwen2VLForConditionalGeneration’s forward?

🐢 No, it has some conditionals:

        if inputs_embeds is None:
            inputs_embeds = self.model.embed_tokens(input_ids)
            if pixel_values is not None:
                pixel_values = pixel_values.type(self.visual.get_dtype())
                image_embeds = self.visual(pixel_values, grid_thw=image_grid_thw)
                image_mask = (input_ids == self.config.image_token_id).unsqueeze(-1).expand_as(inputs_embeds)
                image_embeds = image_embeds.to(inputs_embeds.device, inputs_embeds.dtype)
                inputs_embeds = inputs_embeds.masked_scatter(image_mask, image_embeds)

            if pixel_values_videos is not None:
                pixel_values_videos = pixel_values_videos.type(self.visual.get_dtype())
                video_embeds = self.visual(pixel_values_videos, grid_thw=video_grid_thw)
                video_mask = (input_ids == self.config.video_token_id).unsqueeze(-1).expand_as(inputs_embeds)
                video_embeds = video_embeds.to(inputs_embeds.device, inputs_embeds.dtype)
                inputs_embeds = inputs_embeds.masked_scatter(video_mask, video_embeds)

            if attention_mask is not None:
                attention_mask = attention_mask.to(inputs_embeds.device)

        outputs = self.model(
            input_ids=None,
            position_ids=position_ids,
            attention_mask=attention_mask,
            past_key_values=past_key_values,
            inputs_embeds=inputs_embeds,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )

        hidden_states = outputs[0]
        return hidden_states

🦊 The patcher only modifies the past_key_values_args and moves them to a DynamicCache. There are no other changes for patching.

🐢 But the return types are also different. Qwen2VLForConditionalGeneration.forward returns a Union[Tuple, Qwen2VLCausalLMOutputWithPast], but ColQwen2.forward returns a torch.Tensor.

🐇 How is that torch.Tensor calculated from the return type of Qwen2VLForConditionalGeneration? ❓

🦊 There are multiple return types in Qwen2VLForConditionalGeneration. It’s determined by the return_dict argument.

        if not return_dict:
            output = (logits,) + outputs[1:]
            return (loss,) + output if loss is not None else output

        return Qwen2VLCausalLMOutputWithPast(
            loss=loss,
            logits=logits,
            past_key_values=outputs.past_key_values,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
            rope_deltas=self.rope_deltas,
        )

🐇 What’s that argument in ColQwen2?

🐢 It’s passed from the caller; it’s a kwarg.

🐇 Is there any modification to this in the ONNX patcher?

🐢 Nope.

🐇 Then we can consider it as the default. What’s the default?

🐢 The default is None. Hence it returns (logits,) + outputs[1:].

🐇 Then, logits are the first parameter by default.

🐢 Yes, we can assume this.

🐇 What does ColQwen2 do with these logits?

🦊 By the way, it calls inner_forward with use_cache=False and output_hidden_states=True. What do these change in Qwen2VLForConditionalGeneration?

🐢 These hidden states are actually output. In ColQwen2.forward, the inner_forward call is actually:

        last_hidden_states = self.inner_forward(
            *args, **kwargs, position_ids=position_ids, use_cache=False, output_hidden_states=True
        )  # (batch_size, sequence_length, hidden_size)

and the output is the hidden states.

🦊 It then runs:

        proj = self.custom_text_proj(last_hidden_states)  # (batch_size, sequence_length, dim)

and calculates a set of projections.

🐢 custom_set_proj is defined as:

        self.custom_text_proj = nn.Linear(self.model.config.hidden_size, self.dim)

hence these projections are fully connected layer calculations.

🦊 It returns these after L2 normalization and checks if there are pixel_values to consider.

🐢 In this case, the output from ColQwen2 is this single tensor.

🐇 Ok. Then it’s actually simpler than wrapping up the whole Qwen2VLForConditionalGeneration.

🐢 It looks so, yes. But the example for ColQwen2 also has a post-processing step:

scores = processor.score_multi_vector(query_embeddings, image_embeddings)

🐇 I think we can leave that post-processing for the time being. We only need multi-vector embeddings for FastEmbed.

🐢 Maybe it’s convertible to a forward method that we can use with ONNX.

🐇 Let’s see how this scoring works, then.

🐢 score_multi_vector receives two tensors or tensor lists and compares query vectors with passage vectors. It’s a rather straightforward implementation that requires Torch, but not Transformers.

🐇 In theory, we can also convert this to an ONNX model.

🦊 We can also write a custom model for this.

🐢 It has two for loops for comparisons. Can we convert all of these?

🦊 The loop indices can be seen as dynamic_axes, and it’s possible to convert the whole thing as a Torch model, then use torch.onnx.export just as we do for the model itself.

🐢 I see, but I don’t think that’s what we must do now.

🐇 Yes, let’s skip that for the time being and convert the model itself. We’ll have multivectors for patches and queries at the end.

🐢 So, we’ll keep the processing part, send BatchFeature objects that are output from the processor, and send this to two models: one for images and one for text.

🐇 Yep, that’s the plan. In the end, we’ll have two ONNX models that require BatchFeatures.

🐢 Then, we’ll modify past_key_values in this argument to use DynamicCache.

🐇 Yes, that’s alright. We can start by moving past_key_values_converter to a method in PatchedColQwen2.

🐢 past_key_values wasn’t used much, so I completely removed it. I also began to use processor outputs as dummy input in the exporter. However, when using BatchFeature, we get:

RuntimeError: Only tuples, lists and Variables are supported as JIT inputs/outputs. Dictionaries and strings are also accepted, but their usage is not recommended. Here, received an input of unsupported type: BatchFeature

🐇 So, we can try Dynamo first, I think. If that doesn’t work, we can just collect items as tensors and build up a BatchFeature inside the patcher.

🐢 Let’s try dynamo=True for this first.

🦊 This time, it’s about BatchFeature again, but the error is different: KeyError: 'Indexing with integers is not available when using Python based feature extractors'

🐇 In this case, we can just create BatchFeature inside the patcher.

PatchedColQwen2

class PatchedColQwen2(ColQwen2):
    def forward(self, *args):
        (
            input_ids,
            inputs_embeds,
            attention_mask,
            position_ids,
            *past_key_values_args,
        ) = args
        # Convert past_key_values list to DynamicCache
        if len(past_key_values_args) == 0:
            past_key_values = None
        else:
            past_key_values = DynamicCache()
            for i in range(self.config.num_hidden_layers):
                key = past_key_values_args.pop(0)
                value = past_key_values_args.pop(0)
                past_key_values.update(key_states=key, value_states=value, layer_idx=i)

        breakpoint()
        o = super().forward(
            input_ids=input_ids,
            inputs_embeds=inputs_embeds,
            attention_mask=attention_mask,
            # position_ids=position_ids,
            past_key_values=past_key_values,
        )

        flattened_past_key_values_outputs = {
            "logits": o.logits,
        }
        output_past_key_values: DynamicCache = o.past_key_values
        for i, (key, value) in enumerate(
            zip(output_past_key_values.key_cache, output_past_key_values.value_cache)
        ):
            flattened_past_key_values_outputs[f"present.{i}.key"] = key
            flattened_past_key_values_outputs[f"present.{i}.value"] = value

        return flattened_past_key_values_outputs

🐇 While checking macos-option-as-alt as a potential source of the problem, I spent time adjusting the Ghostty icon. I changed the app icon. This is how I spend my time being extremely productive.

🐢 Weird. The Option as alt key setting is fine.

🐇 Actually, the Alt key works; the reason we cannot make panes larger is that Alt-Right and Alt-Left are not bound to this.

🐢 You may be right, yes. Then, what’s the key to enlarge a pane?

🦊 List all the keys. Check the keys. You should have done this already.

$ tmux list-keys | rg resize

bind-key    -T prefix       >                      display-menu -T "#[align=centre]#{pane_index} (#{pane_id})" -x P -y P "#{?#{m/r:(copy|view)-mode,#{pane_mode}},Go To Top,}" < { send-keys -X history-top } "#{?#{m/r:(copy|view)-mode,#{pane_mode}},Go To Bottom,}" > { send-keys -X history-bottom } '' "#{?mouse_word,Search For #[underscore]#{=/9/...:mouse_word},}" C-r { if-shell -F "#{?#{m/r:(copy|view)-mode,#{pane_mode}},0,1}" "copy-mode -t=" ; send-keys -X -t = search-backward "#{q:mouse_word}" } "#{?mouse_word,Type #[underscore]#{=/9/...:mouse_word},}" C-y { copy-mode -q ; send-keys -l "#{q:mouse_word}" } "#{?mouse_word,Copy #[underscore]#{=/9/...:mouse_word},}" c { copy-mode -q ; set-buffer "#{q:mouse_word}" } "#{?mouse_line,Copy Line,}" l { copy-mode -q ; set-buffer "#{q:mouse_line}" } '' "#{?mouse_hyperlink,Type #[underscore]#{=/9/...:mouse_hyperlink},}" C-h { copy-mode -q ; send-keys -l "#{q:mouse_hyperlink}" } "#{?mouse_hyperlink,Copy #[underscore]#{=/9/...:mouse_hyperlink},}" h { copy-mode -q ; set-buffer "#{q:mouse_hyperlink}" } '' "Horizontal Split" h { split-window -h } "Vertical Split" v { split-window -v } '' "#{?#{>:#{window_panes},1},,-}Swap Up" u { swap-pane -U } "#{?#{>:#{window_panes},1},,-}Swap Down" d { swap-pane -D } "#{?pane_marked_set,,-}Swap Marked" s { swap-pane } '' Kill X { kill-pane } Respawn R { respawn-pane -k } "#{?pane_marked,Unmark,Mark}" m { select-pane -m } "#{?#{>:#{window_panes},1},,-}#{?window_zoomed_flag,Unzoom,Zoom}" z { resize-pane -Z }
bind-key    -T prefix       z                      resize-pane -Z
bind-key -r -T prefix       M-Up                   resize-pane -U 5
bind-key -r -T prefix       M-Down                 resize-pane -D 5
bind-key -r -T prefix       M-Left                 resize-pane -L 5
bind-key -r -T prefix       C-Up                   resize-pane -U
bind-key -r -T prefix       C-Down                 resize-pane -D
bind-key -r -T prefix       C-Left                 resize-pane -L
bind-key -r -T prefix       C-Right                resize-pane -R
bind-key    -T root         MouseDown3Pane         if-shell -F -t = "#{||:#{mouse_any_flag},#{&&:#{pane_in_mode},#{?#{m/r:(copy|view)-mode,#{pane_mode}},0,1}}}" { select-pane -t = ; send-keys -M } { display-menu -T "#[align=centre]#{pane_index} (#{pane_id})" -t = -x M -y M "#{?#{m/r:(copy|view)-mode,#{pane_mode}},Go To Top,}" < { send-keys -X history-top } "#{?#{m/r:(copy|view)-mode,#{pane_mode}},Go To Bottom,}" > { send-keys -X history-bottom } '' "#{?mouse_word,Search For #[underscore]#{=/9/...:mouse_word},}" C-r { if-shell -F "#{?#{m/r:(copy|view)-mode,#{pane_mode}},0,1}" "copy-mode -t=" ; send-keys -X -t = search-backward "#{q:mouse_word}" } "#{?mouse_word,Type #[underscore]#{=/9/...:mouse_word},}" C-y { copy-mode -q ; send-keys -l "#{q:mouse_word}" } "#{?mouse_word,Copy #[underscore]#{=/9/...:mouse_word},}" c { copy-mode -q ; set-buffer "#{q:mouse_word}" } "#{?mouse_line,Copy Line,}" l { copy-mode -q ; set-buffer "#{q:mouse_line}" } '' "#{?mouse_hyperlink,Type #[underscore]#{=/9/...:mouse_hyperlink},}" C-h { copy-mode -q ; send-keys -l "#{q:mouse_hyperlink}" } "#{?mouse_hyperlink,Copy #[underscore]#{=/9/...:mouse_hyperlink},}" h { copy-mode -q ; set-buffer "#{q:mouse_hyperlink}" } '' "Horizontal Split" h { split-window -h } "Vertical Split" v { split-window -v } '' "#{?#{>:#{window_panes},1},,-}Swap Up" u { swap-pane -U } "#{?#{>:#{window_panes},1},,-}Swap Down" d { swap-pane -D } "#{?pane_marked_set,,-}Swap Marked" s { swap-pane } '' Kill X { kill-pane } Respawn R { respawn-pane -k } "#{?pane_marked,Unmark,Mark}" m { select-pane -m } "#{?#{>:#{window_panes},1},,-}#{?window_zoomed_flag,Unzoom,Zoom}" z { resize-pane -Z } }
bind-key    -T root         MouseDrag1Border       resize-pane -M
bind-key    -T root         M-MouseDown3Pane       display-menu -T "#[align=centre]#{pane_index} (#{pane_id})" -t = -x M -y M "#{?#{m/r:(copy|view)-mode,#{pane_mode}},Go To Top,}" < { send-keys -X history-top } "#{?#{m/r:(copy|view)-mode,#{pane_mode}},Go To Bottom,}" > { send-keys -X history-bottom } '' "#{?mouse_word,Search For #[underscore]#{=/9/...:mouse_word},}" C-r { if-shell -F "#{?#{m/r:(copy|view)-mode,#{pane_mode}},0,1}" "copy-mode -t=" ; send-keys -X -t = search-backward "#{q:mouse_word}" } "#{?mouse_word,Type #[underscore]#{=/9/...:mouse_word},}" C-y { copy-mode -q ; send-keys -l "#{q:mouse_word}" } "#{?mouse_word,Copy #[underscore]#{=/9/...:mouse_word},}" c { copy-mode -q ; set-buffer "#{q:mouse_word}" } "#{?mouse_line,Copy Line,}" l { copy-mode -q ; set-buffer "#{q:mouse_line}" } '' "#{?mouse_hyperlink,Type #[underscore]#{=/9/...:mouse_hyperlink},}" C-h { copy-mode -q ; send-keys -l "#{q:mouse_hyperlink}" } "#{?mouse_hyperlink,Copy #[underscore]#{=/9/...:mouse_hyperlink},}" h { copy-mode -q ; set-buffer "#{q:mouse_hyperlink}" } '' "Horizontal Split" h { split-window -h } "Vertical Split" v { split-window -v } '' "#{?#{>:#{window_panes},1},,-}Swap Up" u { swap-pane -U } "#{?#{>:#{window_panes},1},,-}Swap Down" d { swap-pane -D } "#{?pane_marked_set,,-}Swap Marked" s { swap-pane } '' Kill X { kill-pane } Respawn R { respawn-pane -k } "#{?pane_marked,Unmark,Mark}" m { select-pane -m } "#{?#{>:#{window_panes},1},,-}#{?window_zoomed_flag,Unzoom,Zoom}" z { resize-pane -Z }

🐇 resize-pane has keys, but they conflict with C-Right, etc., which are macOS display selection keys. I defined M-S-Right, etc., to resize and M-C-Right, etc., to swap the panes. Defining new keys with M-S…

bind-key -n "M-S-Right" resize-pane -R 10
bind-key -n "M-S-Left" resize-pane -L 10
bind-key -n "M-S-Up" resize-pane -U 10
bind-key -n "M-S-Down" resize-pane -D 10

🐢 We can set keys to swap windows too, maybe to M-C-PageDown, etc. M-PageDown now moves between windows.

🐇 Let’s take a look at the command options:

     swap-window [-d] [-s src-window] [-t dst-window]
                   (alias: swapw)
             This is similar to link-window, except the source and destination windows are swapped.  It is an error if no window exists at src-window.  If -d is given, the new window
             does not become the current window.

             If -s is omitted and a marked pane is present (see select-pane -m), the window containing the marked pane is used rather than the current window.

🐢 The use case will be limited, though.

🐇 Umm, right. Let’s stop procrastinating here.

🐢 It was productive procrastination, though. Now I can resize my tmux panes.

🐇 We can continue working on rclone.

🐢 Umm, ok. Let’s try to focus on adding another storage type.

• ✅ #🌻 ADD rclone storage type (2025-04-24 06:27)

🐢 I think the first option is to run the commands from the command line. We can just use a modified generic storage type without trying to make it fast.

🐇 Let’s make it run first, you say?

🐢 Yes, let’s make it run first and then we can think about making it run fast.

🐇 You’re right!

• ✅ #🌻 ADD generic rclone tests (2025-04-24 06:27)

It’s possible to use the alias remote with a local path.

🐢 Created the PR, waiting for the tests.

🦊 Let’s search if there is a crate to manage rclone. Maybe it will be easier that way.

🐇 It will add another dependency though.

🦊 We can always add a feature flag for this.

🐢 There is a librclone crate that can be used to call rclone commands like https://rclone.org/rc/#supported-commands

🦊 We can test it from the command line perhaps.

rclone rc 
2025/03/15 17:41:09 NOTICE: Failed to rc: failed to list: connection failed: Post "http://localhost:5572/rc/list": dial tcp [::1]:5572: connect: connection refused

🐇 It requires the backend to be running in the background.

🦊 There may be examples in the repository.

🐢 There are none. We can search GH for this crate though.

🦊 It’s also possible to search for dependents in crates.io.

🐢 I think Xvc will be the first dependent of this crate: https://crates.io/crates/librclone/reverse_dependencies

🐇 The following two projects depend on librclone:

• https://github.com/Sh3mm/WarpDrive/tree/master
• https://github.com/hwittenborn/celeste

🐢 Let’s clone Celeste. It uses librclone and looks like it’s a user interface for rclone written in Rust.

🐇 The examples are in celeste/src/rclone.rs.

🐢 Cool. Let’s take a look at how commands are run:

    /// Common function for some of the below command.
    fn common(command: &str, remote_name: &str, path: &str) -> Result<(), RcloneError> {
        let resp = run(
            command,
            &json!({
                "fs": get_remote_name(remote_name),
                "remote": util::strip_slashes(path),
            })
            .to_string(),
        );

        match resp {
            Ok(_) => Ok(()),
            Err(json_str) => Err(serde_json::from_str(&json_str).unwrap()),
        }
    }

All commands are run like librclone::rpc(method, input)) and the commands are like:

    /// make a directory on the remote.
    pub fn mkdir(remote_name: &str, path: &str) -> Result<(), RcloneError> {
        common("operations/mkdir", remote_name, path)
    }

🐇 We have all commands in this file that are relevant to Xvc. Let’s list them here:

• make directory: common("operations/mkdir", remote_name, path)
• delete file: common("operations/delete", remote_name, path)
• remove a dir and all of its contents: common("operations/purge", remote_name, path)
• copy file:

run( "operations/copyfile",
            &json!({
                "srcFs": src_fs,
                "srcRemote": util::strip_slashes(src_remote),
                "dstFs": dst_fs,
                "dstRemote": util::strip_slashes(dst_remote)
            })

and

    /// Copy a file from the local machine to the remote.
    pub fn copy_to_remote(
        local_file: &str,
        remote_name: &str,
        remote_destination: &str,
    ) -> Result<(), RcloneError> {
        copy(
            "/",
            local_file,
            &get_remote_name(remote_name),
            remote_destination,
        )
    }

    /// Copy a file from the remote to the local machine.
    pub fn copy_to_local(
        local_destination: &str,
        remote_name: &str,
        remote_file: &str,
    ) -> Result<(), RcloneError> {
        copy(
            &get_remote_name(remote_name),
            remote_file,
            "/",
            local_destination,
        )
    }

🐇 It looks like that’s all we need. We can organize the commands differently, but these examples are enough to use librclone. It seems rather straightforward.

ghrl | get url | first
https://github.com/iesahin/xvc/actions/runs/13875177382

🐢 We forgot to update the doc tests. Let’s run them again to update storage remove and file untrack commands.

🐇 There are issues with elision.

ghpl

╭───┬──────────────────────┬────────────────────┬──────────────────────────────╮
│ # │     headRefName      │       title        │             url              │
├───┼──────────────────────┼────────────────────┼──────────────────────────────┤
│ 0 │ storage-remove-16674 │ xvc storage remove │ https://github.com/iesahin/x │
│   │                      │                    │ vc/pull/270                  │
╰───┴──────────────────────┴────────────────────┴──────────────────────────────╯

🐢 Tests are passing; we can merge the PR. But our commit hook to check the CHANGELOG doesn’t work. That’s weird; we don’t get any errors when the CHANGELOG is not in the push set.

tmux new-window -c ($env.HOME | path join github.com iesahin xvc.py)  nvim 

🐢 Now we can update the Python bindings as well.

🐇 We forgot to bump package versions. We need to create a checklist for releases.

• ✅ #🌻 CREATE a release checklist (2025-03-25 17:52)

🐢 Let’s check if the latest version is updated on PyPI.

pypi xvc

🐇 Yes, it is.

🐇 What are the options, though? xvc.py is just a wrapper around xvc, and that was the easiest way to get it working. We can list all options manually in the headers as documentation, but it will be harder to maintain.

🐢 Once we start doing it, we’ll find a good way to simplify and shorten it.

🐇 Let’s start to work on xvc file track, then.

🐢 Worked on it a bit, and I decided it’s not worth it at the moment. We need to supply default values for most of the options. Maintaining a separate list of default values may not be feasible; it will be error-prone in a different way.

🐇 It should only be needed to update the dependency versions in Cargo.toml, right?

🐢 Let’s start with that.

🐇 We have interface changes regarding aliases; let’s start using uv for building.

🦊 Added requirements to pyproject.toml by running:

uv add -r requirements.txt

🐢 There is a way if we start from the configuration, not just files. The default configuration is a TOML document. We can make it an XvcConfiguration struct and load and store it with confy.

🐇 We have a cascading set of configurations, but maybe we can start from the struct and serialize/deserialize it on demand.

🐲 I don’t think we need to update xvc-config at the moment. It’s working, and we don’t need to alter its inner workings in the near future.

🐢 I agree. We can use the config crate when we need to update and have enough time to work on this.

🐢 Yes, but what will T be? Although store structures have Serde implementations, they are not particularly useful for this.

🐇 We can have output types, named like XvcFileListOutput, that will be converted to strings with serialization.

🦊 We do something similar in xvc pipeline export and import commands. We use XvcPipelineSchema and XvcStepSchema just for the import and export commands. We’ll write similar structs for all JSON output and will use Serde to convert these to strings.

🐢 Unlike the import and export commands, we have optional fields in the output, though. I don’t want content digests to appear in JSON output if they are not required.

🐇 Let’s search for optional fields in Serde.

🦊 There is a crate for optional fields.

🐢 We don’t need another crate for this. Serde has the skip_serializing_if attribute for fields. We can add Option::is_none as a method to these to skip outputting None fields. All those fields, in this case, will be optional.

🐇 This is fine. We already use structs to format the xvc file list output. We can just use them to output JSON.

🐇 https://www.getzola.org/themes/pico/ is an option, but I don’t like its header.

🦊 Minimal Dark from the same author looks better: https://kuznetsov17.github.io/minimal-dark/notes/note1/

🐇 https://www.getzola.org/themes/no-style-please/ is also an option.

🦊 https://halve-z.netlify.app/posts/information/ looks interesting, but there is too much screen estate for the left bar.

🐢 Let’s start by running the site locally first.

🐇 There are errors in the configuration. That’s weird, but let’s fix these.

🐢 Fixed errors. These are probably related to a newer version. Can we take a look at netlify.toml to see if it downloads the same version?

🐇 There are breaking changes in Zola 0.19. Let’s update and push the Netlify config to see the results.

🐢 We need to add language support for Nushell to prevent warnings. Take a look at how to add a syntax file to Zola.

🐇 I added

extra_syntaxes_and_themes = ["syntaxes"]

to the [markdown] section and added a syntaxes/nushell.sublime-syntax file copied from https://github.com/kurokirasama/nushell_sublime_syntax. I’m getting:

Error: Reason: Error while compiling regex '\b(?x: 7z | ?
...
s-to-gdrive | usage | ver | verify | weather | wget-all | which-cd | wifi-info | wifi-pass | xls2csv | ydx | yt-api | ytcli | ytm | z | zi)\b'
Oniguruma error: target of repeat operator is not specified

🐢 The syntax highlighter may be a bit buggy. Let’s try to fix this if it’s a one-off.

🐇 Found the bug. There is a ? in the regex that causes it to fail. Now it compiles, and Nushell blocks are colored.

🐢 Cool. Let’s fix the other warnings now. shell and console are not recognized, it looks like.

🐇 There is only Bash listed in https://www.getzola.org/documentation/content/syntax-highlighting/.

🦁 What do you think about migrating to mdBook? We already maintain mdBook for XVC; what about just moving the site to mdBook?

🐢 I thought about this before, and the only downside is the lack of an RSS feed.

🦊 I found this: https://github.com/theowenyoung/mdbook-rss

🐢 Now, this changes everything. We can even move the nedriy.at site to mdBook in this case.

🐇 Then, let’s start working on this. The site will be a technical book site in this case.

🐢 Does mdBook support Nushell syntax?

🐇 Nushell is not in the listed languages. mdBook uses highlight.js, and in its listed languages, we don’t find Nu either.

🐢 We already added Nu support to Zola, and we can just change the theme. This is a blocker in my opinion.

🐇 I searched for Nushell highlight.js support, and nothing appears. I think we can just postpone until Nu has more support on this front.

🐢 Yes. Let’s first try this change in the non-technical blog, and we can come back to this issue. Now, we’ll update console and shell to bash, I think.

🐇 Replaced shell and console with bash.

🐢 There is a file for ggplot that has warnings from earlier incarnations. We also lack syntax highlighters for Vim and Tmux.

🐇 There is a sublime-syntax file for Tmux at https://raw.githubusercontent.com/gerardroche/sublime-tmux/refs/heads/master/Tmux.sublime-syntax, but do we need it for a single file?

🐢 Let’s set it to plain text.

🐇 Now we only have Mermaid warnings left.

🐢 There should be a diagram at https://emresahin.net/developing-a-gitignore-crate/, but it doesn’t show up. We need a shortcode to show these, like the YouTube shortcode. Now let’s get back to theme selection.

🐇 I tested Karzok, but it doesn’t have category and tags support.

🐢 And I tested https://github.com/micahkepe/radion, but the best so far is the apollo theme. I’m struggling to modify the index page, though. I forgot that I modified the theme’s index.html file. I have content in /content/_index.md and a modified index.html in /themes/anemone/templates/index.html to show the content, tags, categories, etc. It should be fixed now.

🐇 Ah, cool. Can we clean the recent duplicate pages now?

🐢 Yeah, let’s take a look.

🐇 What about adding at least command completions for Nushell?

🐢 Let’s read a bit about clap_complete_nushell.

🦊 There seems to be a nu-complete command. Let’s check its documentation.

🐇 Nothing was found, and Kagi doesn’t help much either.

🐢 There is a completions document for Nushell: https://www.nushell.sh/book/custom_completions.html

🐇 There is a tool called carapace to provide completions across shells.

🐢 Its documentation is thin, and I’m not sure if it supports dynamic completions out of the box. I believe instead of adding a carapace setup, we can just write a Nushell completion script that will use JSON output from the commands and add some (maybe hidden) utility commands to support it.

🐇 There are a set of example scripts in the Nushell repo: https://github.com/nushell/nu_scripts/tree/main/custom-completions

🐢 The reason I want to write custom completions for Nushell is that it will be an exercise for the scripting language. gh completions are not as scary as a Bash script.

🐇 git completions are a better example for XVC. They simply run git whenever necessary. We can start from a static completions command and update this with dynamic completions manually. It will teach a lot.

🐢 I forked the nu_scripts repo and will add XVC completions script there.

🐇 Then let’s begin by adding Nushell static completions. Shall we add a command for this?

🦊 Reviving the completion command we removed in 0.6.13?

🐢 We shouldn’t list it. We can make a _comp subcommand for the time being and generate and distribute completions in the repository. When clap_complete_nushell has the feature parity to provide dynamic completions, we can remove these commands.

🐇 What will we use this for other than generating completions?

🐢 Maybe dynamic completions can call this as well.

🐇 Added Nushell static completions to be output using xvc _comp generate-nushell. Let’s bump up the version to 0.6.15.

cargo set-version 0.6.15-alpha.1
   Upgrading xvc from 0.6.14 to 0.6.15-alpha.1
...

🐢 I noticed we forgot a line in the CLI command handler that asserts xvc_root_opt.is_some(), and this fails when we run xvc outside of repositories. We need to release this version quickly.

🐇 Oops, now, ok, let’s write a static Nushell generator and just release quickly.

🦊 Generating completions with

xvc comp generate-nushell

🐢 Completion command is run with comp instead of _comp. Should we rename it?

🐇 Renamed it to _comp. It’s not hidden, but at least we can be sure that it won’t be misunderstood as a common command.

🐢 Bumping the version again. Now let’s source the generated script and test it.

cargo set-version 0.6.15-alpha.2
   Upgrading xvc from 0.6.15-alpha.1 to 0.6.15-alpha.2
...

🦊 Yep, it works. We now have completions for Nushell.

🐢 Let’s update the completions documentation.

🐇 Done. Now, let’s take a look at CI and see what fails.

ghrl | first
╭──────────────┬─────────────────────────────────────────────────────────╮
│ conclusion   │ success                                                 │
│ displayTitle │ Add Nushell completions                                 │
│ headBranch   │ nushell-completions                                     │
│ url          │ https://github.com/iesahin/xvc/actions/runs/13070200765 │
╰──────────────┴─────────────────────────────────────────────────────────╯

🐢 It fails because of coverage, not the tests. Codecov says the new code isn’t tested.

🐇 The added xvc _comp command isn’t tested. We can add a test running those lines and testing if the command outputs a completion script.

🐢 We have a test for completions. We can add a test that runs the lines.

🐇 Added a test and bumping up the version.

cargo set-version 0.6.15-alpha.3
   Upgrading xvc from 0.6.15-alpha.2 to 0.6.15-alpha.3
...

🦊 We can add some more coverage while waiting for the tests.

🐇 XvcOutputLine implementation seems to have no tests. It’s weird because we use these everywhere.

🐢 I’m not sure we use this particular implementation; we just use XvcOutputLine::Info(s), not XvcOutputLine::info(s) anywhere. We can delete these methods actually.

🐇 We’ll add JSON output via this particular struct. Can we refactor these to use formatting for JSON, for example? Or use these to output JSON?

🦊 We can add a formatter to XvcOutputLine to output structures.

🐢 The enum is now defined as:

#[derive(Clone, Debug)]
pub enum XvcOutputLine {
    /// The output that we should be reporting to user
    Output(String),
    /// For informational messages
    Info(String),
    /// For debug output to show the internals of Xvc
    Debug(String),
    /// Warnings that are against some usual workflows
    Warn(String),
    /// Errors that interrupts a workflow but may be recoverable
    Error(String),
    /// Panics that interrupts the workflow and ends the program
    /// Note that this doesn't call panic! automatically
    Panic(String),
    /// Progress bar ticks.
    /// Self::Info is also used for Tick(1)
    Tick(usize),
}

Here, these fields can also have a formatter that will render the string in a particular format. For example, the output can be

XvcOutputLine::Output(XvcJsonFormatter, String)

🐇 I’m not sure this is a good idea. Output already specifies this string as output. We can have a wrapper instead, like,

struct XvcJsonOutput(Format<XvcStructuredOutput>, XvcStructuredOutput)

and we can use the supplied format to render XvcStructuredOutput to an output line with XvcOutputLine::Output. If we don’t provide output as structured, it will be too much error-prone work to convert the current outputs to structured.

🦊 The transition will also be gradual. We may not need structured output for most of the commands. We can start with xvc file list and convert others as we go.

🐢 This is sensible. By the way, coverage still didn’t increase. There may be something going on with Codecov or running the test.

ghrl | first
╭──────────────┬─────────────────────────────────────────────────────────╮
│ conclusion   │ success                                                 │
│ displayTitle │ Add Nushell completions                                 │
│ headBranch   │ nushell-completions                                     │
│ url          │ https://github.com/iesahin/xvc/actions/runs/13087431038 │
╰──────────────┴─────────────────────────────────────────────────────────╯

🐇 Let’s run the test:

cargo test -p xvc --test test_completions
...
test test_completions ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.80s

🐢 Can we make sure the output is a Nushell script and not an error message?

🐇 Let’s print it out.

🐢 It looks like when the COMPLETE environment variable is set, it never calls _comp subcommand and never calls those lines.

cargo set-version 0.6.15-alpha.4
   Upgrading xvc from 0.6.15-alpha.3 to 0.6.15-alpha.4
...

🐢 Let’s make a release for 0.6.15. Coverage is OK now.

cargo set-version 0.6.15
   Upgrading xvc from 0.6.15-alpha.4 to 0.6.15
...

gh pr merge --squash --body $"(open CHANGELOG.md | lines | skip 2 | take 5)" --subject "Add static nushell completions"

🐇 Merged the PR.

🐢 Releases should appear in a few minutes.

🐇 We need to tag the merge commit for this.

🐢 Oh, yep. AFAIK Lazygit doesn’t have something for git push --tags. Let’s push from the CLI.

git push --tags
You are on the main branch. Skipping CHANGELOG.md check.
To github.com:iesahin/xvc
 * [new tag]         v0.6.15 -> v0.6.15

🦊 These commands, especially tables, are not rendered correctly on the web. We need to change the theme, I think.

🐇 There are two errors. One resulted from forgetting sudo when installing dependencies. The other was the incorrect name for the OpenSSL library. It should be libssl-dev instead of openssl-dev. Maybe we can link the command list in the docs.

🐢 Oops, yes, we should at least keep that in mind.

🐇 I’m checking the most popular crates. getrandom retrieves a random number from the system. It looks much lighter than the rand crate.

🦊 We can use once_cell to initialize XvcEntityCounter. We currently use Once for this purpose.

🐇 I don’t think it will provide any better features.

🦊 For that case, yes, no better features. But the interface is something like:

impl<T> OnceCell<T> {
    const fn new() -> OnceCell<T> { ... }
    fn set(&self, value: T) -> Result<(), T> { ... }
    fn get(&self) -> Option<&T> { ... }
}

And this makes, for example, working with XvcRoot much easier. We are passing Arc<RwLock<XvcRootInner>>> everywhere. This is a heavy price when we only use it in a read-only manner. We can prevent most of these, when we use a read lock, by using OnceCell.

🐇 XvcConfig can benefit from this as well. We don’t update the config during runs.

🐢 Why do we want to assign it, though? We currently have a config field in XvcRootInner, and we get a reference to it with the config() method.

🐇 Ok. Let’s skip this for now. No need to worry before measuring the performance impact.

$ ghrl | first
╭──────────────┬─────────────────────────────────────────────────────────╮
│ conclusion   │ failure                                                 │
│ displayTitle │ Add CLI completions                                     │
│ headBranch   │ clap-complete-16608                                     │
│ url          │ https://github.com/iesahin/xvc/actions/runs/13007688178 │
╰──────────────┴─────────────────────────────────────────────────────────╯

🐢 Although I turned off most of the watches, logs are still so large that it’s not possible to view them from the interface. Downloaded the log archive.

🦊 We can have different GitHub Actions steps for each test. Claude can help write such a repeating set of steps.

🐇 We can at least separate z_test_docs to see if integration tests or that fails.

🐢 We’ll have to add caching for test artifacts to upload them for coverage. I’m not sure we really need to add that complexity to the process just to avoid downloading the logs.

🐇 We can actually move all to Xvc. We need a GitHub Action to run an Xvc pipeline.

🐢 Eventually yes, we should move our testing to Xvc itself. Now, the logs show that there are differences in z_test_docs actually.

🐇 When I run the command below, it passes. We may have a different config in GitHub Actions.

$ XVC_TRYCMD_TESTS=storage,file,pipeline,core,start TRYCMD=overwrite rws cargo test --features test-ci -p xvc --test z_test_docs
test z_doc_tests ... ok
...
test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 15.34s

🐢 We don’t have an XVC_TRYCMD_TESTS=storage,file,pipeline,core,start definition in GitHub Actions; let’s add it, bump the version, and try again.

$ cargo set-version "0.6.14-alpha.10"
   Upgrading xvc from 0.6.14-alpha.9 to 0.6.14-alpha.10
...

🦊 We could also update run-tests.zsh to get a quick response.

🐢 Yep, let’s do that as well.

🐇 Dev tests pass, but there are differences in the documents still. For some reason, the local run-tests.zsh doesn’t update xvc/book/src/ref/xvc-storage.md. It only has the help text as a reference, but it’s not updated with the aliases, and it breaks the CI. This is weird but a small issue. Fixed it manually.

$ ghrl | first 2
╭───┬────────────┬───────────────────┬────────────────────┬────────────────────╮
│ # │ conclusion │   displayTitle    │     headBranch     │        url         │
├───┼────────────┼───────────────────┼────────────────────┼────────────────────┤
│ 0 │            │ Add CLI           │ clap-complete-1660 │ https://github.com │
│   │            │ completions       │ 8                  │ /iesahin/xvc/actio │
│   │            │                   │                    │ ns/runs/1302770108 │
│   │            │                   │                    │ 2                  │
│ 1 │ success    │ Add CLI           │ clap-complete-1660 │ https://github.com │
│   │            │ completions       │ 8                  │ /iesahin/xvc/actio │
│   │            │                   │                    │ ns/runs/1302755462 │
│   │            │                   │                    │ 2                  │
╰───┴────────────┴───────────────────┴────────────────────┴────────────────────╯

🐢 The earlier one has passed. It looks like we’re ready to merge. Let’s update the CHANGELOG.md for release.

🐇 Setting the release version:

cargo set-version "0.6.14"
   Upgrading xvc from 0.6.14-alpha.10 to 0.6.14
...

🐢 Let’s check the CI

$ ghrl | first 2

╭───┬────────────┬───────────────────┬────────────────────┬────────────────────╮
│ # │ conclusion │   displayTitle    │     headBranch     │        url         │
├───┼────────────┼───────────────────┼────────────────────┼────────────────────┤
│ 0 │ success    │ Add CLI           │ clap-complete-1660 │ https://github.com │
│   │            │ completions       │ 8                  │ /iesahin/xvc/actio │
│   │            │                   │                    │ ns/runs/1302770108 │
│   │            │                   │                    │ 2                  │
│ 1 │ success    │ Add CLI           │ clap-complete-1660 │ https://github.com │
│   │            │ completions       │ 8                  │ /iesahin/xvc/actio │
│   │            │                   │                    │ ns/runs/1302755462 │
│   │            │                   │                    │ 2                  │
╰───┴────────────┴───────────────────┴────────────────────┴────────────────────╯

🐇 And merge:

$ ghpM --body $"(open CHANGELOG.md | lines | skip 2 | take 7)" --subject "Add completions" --squash

🐢 Tagged main and pushed. Packages should be built in a few minutes.

🐢 Instead, we can make it test with environment variables for each of these shells.

🐇 It looks like the test fails with wait_status.

cargo test -p xvc --test test_completions
...
failures:

---- test_completions stdout ----
...
ExitStatus(unix_wait_status(512))

thread 'test_completions' panicked at lib/tests/common/mod.rs:46:5:
Command failed: Command { cmd: "/Users/iex/github.com/iesahin/xvc/target/debug/xvc", stdin: None, timeout: None }
...

🐢 The failure is in the line:

    let mut cmd = Command::cargo_bin("xvc").unwrap();

So it probably waits for user input to complete, but it doesn’t have any and cannot complete it, so it returns an error.

🦊 Let’s comment the earlier test out and check if other tests work.

🐢 Yes, they fail the same. When there are no commands to run, xvc returns an error. Maybe we can change this behavior or handle the error in the Command::cargo_bin line above.

🐇 I checked the code and we don’t handle the “no arguments” case anywhere. It looks like the behavior to return an error code is inherited from clap.

🐢 Updated the error handling code to report a more descriptive message from the source.

cargo test -p xvc --test test_completions
...
failures:

---- test_completions stdout ----
Output { status: ExitStatus(unix_wait_status(25856)), stdout: "", stderr: "\nthread 'main' panicked at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_builder-4.5.20/src/builder/debug_asserts.rs:341:13:\nCommand pipeline: command `export` alias `l` is duplicated\nstack backtrace:\n   0: rust_begin_unwind\n             at /rustc/b1a7dfb91106018f47ed9dc9b27aee1977682868/library/std/src/panicking.rs:692:5\n   1: core::panicking::panic_fmt\n             at /rustc/b1a7dfb91106018f47ed9dc9b27aee1977682868/library/core/src/panicking.rs:75:14\n   2: clap_builder::builder::debug_asserts::assert_app\n             at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_builder-4.5.20/src/builder/debug_asserts.rs:341:13\n   3: clap_builder::builder::command::Command::_build_self\n             at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_builder-4.5.20/src/builder/command.rs:4173:13\n   4: clap_builder::builder::command::Command::_build_recursive\n             at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_builder-4.5.20/src/builder/command.rs:4076:9\n   5: clap_builder::builder::command::Command::_build_recursive\n             at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_builder-4.5.20/src/builder/command.rs:4078:13\n   6: clap_builder::builder::command::Command::build\n             at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_builder-4.5.20/src/builder/command.rs:4071:9\n   7: clap_complete::env::CompleteEnv<F>::try_complete_\n             at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_complete-4.5.40/src/env/mod.rs:229:9\n   8: clap_complete::env::CompleteEnv<F>::try_complete\n             at /Users/iex/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/clap_complete-4.5.40/src/env/mod.rs:210:9\n   9: xvc::main\n             at ./src/main.rs:18:30\n  10: core::ops::function::FnOnce::call_once\n             at... [truncated]
ExitStatus(unix_wait_status(25856))

...

error: test failed, to rerun pass `-p xvc --test test_completions`

🐢 The real issue is when we assert the successful run, at this line:

            assert!(output.status.success(), "Command failed: {:?}", prepared);

🐇 If you look at the error message carefully, you’ll see that this is a different error. We added an alias to xvc pipeline export with l, which is a duplicate.

🐢 Oops, yeah.

cargo test -p xvc --test test_completions
...
test test_completions ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.81s

🐇 There are some warnings in compilation. Let’s fix these.

cargo build
   Compiling xvc-storage v0.6.14-alpha.8 (/Users/iex/github.com/iesahin/xvc/storage)
   Compiling xvc-file v0.6.14-alpha.8 (/Users/iex/github.com/iesahin/xvc/file)
   Compiling xvc-pipeline v0.6.14-alpha.8 (/Users/iex/github.com/iesahin/xvc/pipeline)
   Compiling xvc v0.6.14-alpha.8 (/Users/iex/github.com/iesahin/xvc/lib)
    Finished `dev` profile [unoptimized + debuginfo] target(s) in 4.28s

🐢 Finished the build without warnings. Now we can run the whole test suite again.

🐇 Tests are running fine. Updated some error messages and types, and doc tests also pass. There is an issue with the Rsync storage ref.

🐢 It looks like we try to elide output with [...], while the proper format is [..].

🐇 Replaced them with ... that will elide multiple lines.

🐢 Pushed changes to the server.

🐇 We can just complete the completions for this version and add JSON output for all commands to use in nushell. In the meantime, dynamic completions support for nushell is likely to be added.

🐢 Umm, yep. Then we’ll continue to run tests with zsh for the current version.

🐇 run-tests is a zsh file, and we will continue to use it. We can switch to nushell eventually as we add more compatibility with it, but let’s not do this now.

🐢 Where were we the last time for completions?

🐇 We added xvc_path_completers. Let’s check the TODO list now.

🐢 We still need tracked_targets completions in a few places.

🐇 Added those.

🐢 Now we have some strum completers. Let’s finish these up as well.

🐇 Ok. They are done as well.

🐢 Now, let’s start adding a storage_identifier completer. It should read all XvcStorage records and list their names.

🐇 Added a storage_identifier completer. Let’s add it to all places where we use storage_identifiers.

🐢 Now, we need completers for pipeline and step names. These are straightforward.

🐇 Added these. Let’s fill up where they are needed.

🐢 I want to skip some of the fine-grained completions in this version. We can have a specific completer for --params options for files and params inside, for example. Or a special completer for directories.

🐇 Let’s not forget these. Reading YAML files and extracting hyperparameter keys for the prompt would be a really good feature for the user.

🐢 I agree. These are good features in general, but we need to ship the current version as soon as possible.

🐇 Let’s check the TODO comments once more.

$ rg 'TODO:' 
...
- ✅ pipeline/src/pipeline/api/update.rs:    /// TODO: Add a repository_dirs completer (11:08)
- ✅ file/src/remove/mod.rs:    /// TODO: Add a storage_identifier completer (11:08)
- ✅ file/src/bring/mod.rs:    /// TODO: Add a storage_identifier completer (11:08)
...

🦊 We can use xvc_path_completer for the tracked directory completer for the time being. For pipeline update, we can use a ValueHint instead. It’s not necessary to use a custom completion.

🐢 For the xvc file copy destination, we can have a file or a directory that we track and is not available, or we don’t track and is available. It’s a similar situation to xvc_path_completer, but we also need to check the local paths. It requires some more care.

🐇 Now we can begin to run the tests.

🐢 Is this building?

🐇 It should.

🐢 We also have one bug. xvc shouldn’t print help text when the COMPLETE environment variable is set.

🐇 Umm, yeah, that’s a blocker.

🐢 Fixed it. We can bump the version and push the changes.

🐇 It was one of those nice days. I have that feeling in my head that I used it to its full potential.

🐢 So we can do some light work today.

🐇 Like completions?

🐢 Yep, let’s fill in some missing pieces there.

🐇 Let’s see what those missing pieces are.

🐢 We need to clean up the code comments we copied from jj.

🦊 Cleaned up quite a bit and moved everything to main. It’s a single function call.

🐢 I think we completed all strum-related completions. Now we need to discuss store completions a bit.

🐇 They can’t all have the same function like we used for strum-based enums. They will look up different parts of components. For XvcPath, it will be the inner RelativePathBuf; for StorageIdentifier, that will be the names of all recorded storage names and GUIDs. The field we’re going to complete will change from case to case.

🦊 They can depend on the same machinery. It will detect .xvc, load a store, and return a field from the component. The first two are the same for all.

🐢 Yep. Let’s write a few TODOs in the code, then.

🦊 One thing to discuss is whether we should load all config and root. It may be a bit time-consuming. Just loading the stores should be enough.

🐢 I think so. ECS doesn’t depend on XvcRoot, and we can load stores without loading XvcRoot. We have convenience functions in XvcRoot, but they are for convenience. There is nothing that prevents us from loading stores without loading the root.

🦊 Let’s take a look at XvcRoot’s loading of these.

🐢 Tomorrow, hopefully.

🐇 Starting with completions is better. One question I have is whether the completions work for subcommands as usual, like in the static completions.

🐢 Yes, they work. We need to add:

source <(COMPLETE=zsh xvc)

to .zshrc, though.

🐇 It looks like we can drop the xvc completions command. We’re just checking the COMPLETE environment variable, and there’s no need for a separate command in this case.

🐢 Yes, let’s remove that.

🐇 Maybe we can employ a completions module for all completion-related functionality. It’s a separate thing, you know.

🐢 Which completions do we need?

🐇 We can just mark them with TODOs now.

🐢 Yes, let’s check where we need completions and what.

🐇 I noticed we’re repeating options in XvcFileCLI, and some of the options are missing, e.g., --from-ref and --to-branch.

🐢 Because it can compile into a different binary, and global options don’t work that way. Maybe we can move all these binaries into different files under lib. We can have feature flags to turn off certain features and compile different binaries.

🦊 Let’s search for it; I haven’t seen it before: building different binaries with feature flags using cargo.

🐢 There is no clear-cut solution, but it looks like we can move all binaries to the root with the required-features field.

🐲 We can postpone this to another release.

🐇 Yes, let’s not spend time on this now.

🐢 Should we try making pipeline_name a global option? It’s repeating everywhere.

🐇 That will make it easier to maintain.

🐢 We’ll have to pass pipeline_name to subcommands, though.

🐇 I think we can have a set of global options that we can pass. For the time being, that’s only the pipeline_name.

🐢 Okay. Let’s do this.

🐇 A similar option is the step name for pipeline steps.

🐢 Dependencies will need a revamp in the next version anyway. So let’s keep it for now.

🐲 Also, the semantics of step-name are different for these commands. step new interprets it as a new name, while step dependency interprets it as an existing name. The first can be renamed to --name, and the second can be --to, as one of its aliases suggests.

🐢 Yes, let’s keep it for now, and we’ll continue to work on others.

🦊 I want to post a comment in the clap discussions.

🐢 Wrote it. Now let’s build it after changing the pipeline_name.

🐇 It compiles. Should we test it?

🐢 I think we’ll test after all this completion work is done.

🐇 We did most of the strum-related completion.

🐢 Didn’t test them yet, though.

🐇 This is Rust. It will work if it compiles, and it compiles.

🐇 The examples are only found in the tests: https://github.com/clap-rs/clap/blob/master/clap_complete/tests/testsuite/engine.rs#L604

and we can try to create those random strings this way.

🦊 The issue is that we may not be able to use the derive method to add these custom completers. All examples are using the builder API.

🐢 We should be able to obtain the end result from derives and modify them to use these custom completers, but let’s try to use add = ArgValueCompleter first.

🐇 It looks like we have another problem. We need to add a rust-toolchain.toml file to the project, you know, to keep it on the stable channel.

🐢 Ah, yeah, we now use multiple channels. Let’s do that.

🐇 I found an example, actually:

#[derive(Debug, Parser)]
struct Cli {
    #[arg(long, add = ArgValueCompleter::new(custom_completer))]
    custom: Option<String>,
}

🐢 I’m trying to add this to the --from-ref option of Xvc. It’s not running anything. The custom completer should return a random string, but it doesn’t.

🦊 Maybe test with a constant first.

🐢 It didn’t work that way either.

🐇 Let’s do a cargo clean.

🐢 Nothing has changed.

🐇 Let’s take a look at the output of xvc completions:

xvc completions
...
'(--skip-git)--from-ref=[Checkout the given Git reference (branch, tag, commit etc.) before performing the Xvc operation. This runs \`git checkout <given-value>\` before running the command]:FROM_REF:_default' \
...

🐢 This is the only place --from-ref is mentioned and, as far as I can see, there is nothing that calls a dynamic command here.

🐇 Umm, right. There’s something weird here. Maybe we lack the ArgExt trait or something.

🐢 It should be turned on by clap-complete with its unstable-dynamic feature, but who knows.

🐇 It didn’t work.

🐢 ArgExt is available, but it isn’t being used.

🐇 I think I found the answer in https://jj-vcs.github.io/jj/latest/install-and-setup/#command-line-completion:

source <(COMPLETE=zsh xvc)

is the command we should use.

🐢 It doesn’t produce a completion script, though.

🐇 We have this in jj/cli/src/cli_util.rs:

        if env::var_os("COMPLETE").is_some() {
            return handle_shell_completion(ui, &self.app, &config, &cwd);
        }

🐢 I think the whole handle_shell_completion set is written by them. I’d like to see what jj with COMPLETE=zsh outputs.

🐇 Building it to see now.

🐢 As expected, it calls a function:

❯ COMPLETE=zsh target/debug/jj
#compdef jj
function _clap_dynamic_completer_jj() {
    local _CLAP_COMPLETE_INDEX=$(expr $CURRENT - 1)
    local _CLAP_IFS=$'\n'

    local completions=("${(@f)$( \
        _CLAP_IFS="$_CLAP_IFS" \
        _CLAP_COMPLETE_INDEX="$_CLAP_COMPLETE_INDEX" \
        COMPLETE="zsh" \
        /Users/iex/github.com/etc/jj/target/debug/jj -- ${words} 2>/dev/null \
    )}")

    if [[ -n $completions ]]; then
        _describe 'values' completions
    fi
}

compdef _clap_dynamic_completer_jj jj

🐇 Now we should make it the same; the shell must call xvc to get the completions. That’s how all these will work.

🐢 Our goal now is to make a random string output from the --from-ref completion.

🐇 The plan is to make completions work as quickly as possible.

🐢 We have done it. 🎉🥳

🐇 Cool. Now we need to fill up all those completion methods.

🦊 Yeah. We can start with basic ones and move from there.

🐢 There will probably be architectural changes as well. We cannot just run xvc functions directly. We need to run internal functions, but not through commands. At that point, when the user hits tab, we don’t know which command to run.

🐇 Maybe it’s time to move to a Git library. “What is the best Git library for Rust?”

🐲 Should we use a Git library?

🐢 It looks like Gitoxide is the default Rust way to interact with Git repositories. We can keep our way of using the Git binary and use this as an experimental way to learn.

🦊 Yep. That’s a good idea. We can include it in the lib for the time being and start to use it in completions.

🐲 We can work on completions or the GUI.

🐢 Completions are similar to the Homebrew work. It’s in another ecosystem and feels boring.

🐲 We can try to make it in Rust: “Writing shell completions in Rust.”

🐢 There are two solutions: One is clap_complete (https://docs.rs/clap_complete/latest/clap_complete/) and the other is shell_completion (https://github.com/JoshMcguigan/shell_completion). The latter is very new but may be what we need: writing completions only with Rust.

🐲 The shell_completion crate is very bare-bones. It doesn’t have anything, actually. https://github.com/JoshMcguigan/shell_completion/issues/1

🐇 The feature that we need is dynamic completion. We’ll write something similar to ArgValueCompleter for this.

🦊 I think we can just start a new branch.

🐲 We have a branch for Homebrew; should we merge it?

🐢 I think, yes, we can merge it. We’ll fix it if it breaks. It’s a separate workflow file anyway.

🦊 Created the PR.

ghpl
264	add brew tap	add-brew-tap	OPEN	2025-01-03T05:14:45Z

🐢 Merged it.

🐇 Now let’s create a new branch and add clap-complete to Cargo.toml.

🐲 There are some packages that we need to take a look at. thiserror now has v2.

🐢 Upgraded packages and compiled. It works.

🐲 Let’s update the version.

🐢 Done. Now we can add clap-complete.

🦊 Added with the unstable-dynamic command. Can we build it again?

🐢 Built it. Now we can add a completions subcommand to Xvc.

🐲 Is this the right name for this?

🐢 It will output a shell script that we can source in the shell.

🐲 Okay, let’s add the subcommand now.

🐢 The example doesn’t work with the clap builder.

🐇 Let’s search for it.

🐢 I think we’re extending ourselves a bit when trying to add all dynamic features of completion at once. We can just start by adding completion to xvc-test-helper.

🐇 Yep, that’s a better idea. We can have many more examples, and it’s certainly more straightforward.

🐢 One thing I noticed when working yesterday is that using auto-save was actually preventing many of these swap errors.

🐲 It was a bit slow. Can we take a look at a few others?

🐢 https://github.com/okuuva/auto-save.nvim seems a bit more polished.

🐇 Installed it, but haven’t seen an effect yet.

🐢 We may need to restart this.

🐲 Now, we can get into the test-helper completions.

🐢 Done. Added these quickly as you thought. Now adding this to main seems much easier.

🦊 That was a nice approach. Let’s dive into adding this to main.

🐇 Should we go with a separate command or just an option?

🐲 Our commands have distinct initial letters, allowing them to be used with just those letters. Adding another top-level command for this will make c useless.

🐢 I don’t think that’s a valid concern. We can create aliases for all commands and completions shouldn’t have to have an alias. But I agree that completions should not be a top-level command. It will be run once for installation at most.

🦊 I agree. Let’s call it --completions. It will be run as xvc --completions zsh and will print out the completions.

🐢 Okay. Let’s do this.

🦊 No errors left. Let’s install this version.

🐢 We get:

error: 'xvc' requires a subcommand but one was not provided
  [subcommands: file, init, pipeline, storage, root, check-ignore, aliases, help]

Usage: xvc [OPTIONS] <COMMAND>

For more information, try '--help'.

🦊 Umm, okay. I think we don’t have an option to create a command like that. Then we’ll have to print completions with a subcommand.

🐲 The above discussion is now moot. ❌

🐢 We can try to push, but probably it’s not worth it.

🐲 Let’s not lose time on this. I think we can remove the aliases command and replace it with completions, and add single-letter aliases to subcommands.

🐢 We can have an option in the completions command to print aliases. That will work.

🦊 I don’t think people will use the aliases command if we have single-letter command aliases.

🐢 You may be right. No need to try to maintain that at this time.

🐢 Good morning, and I’m fed up with these blink.cmp errors, you know.

🐇 Reinstalling blink works. It’s interesting to rely on such unreliable software.

🐢 Oh, yeah. It’s interesting. Where were we yesterday?

🐲 We decided to rename the aliases command to completions.

🐇 And waiting for rust-analyzer to complete its analysis.

🐢 Uh, yeah. I see.

🦊 The code itself is very short, actually.

    if let Some(shell) = cli_opts.completions {
        let mut cmd = XvcCLI::command();
        generate(shell, &mut cmd, "xvc", &mut io::stdout());
        return Ok(None);
    }

🐲 We’ll use the output! macro instead of writing to io::stdout(), right?

🐢 Yes, we can rely on the usual output system. It will be slower to create an output thread but shouldn’t matter for outputting a shell script.

🐇 It takes a while for rust-analyzer to scan all the directories, it looks like. When I close the project, it just cannot reload it immediately.

🐢 We can use cargo clean from time to time.

🦊 rust-analyzer finished scanning; let’s try to rename AliasesCLI.

🐲 Let’s keep these here; maybe we’ll need them in our scripts:

# Standard Xvc command aliases for longer commands.
alias xls='xvc file list'
alias pvc='xvc pipeline'
alias fvc='xvc file'
alias xvcf='xvc file'
alias xvcft='xvc file track'
alias xvcfl='xvc file list'
alias xvcfs='xvc file send'
alias xvcfb='xvc file bring'
alias xvcfh='xvc file hash'
alias xvcfco='xvc file checkout'
alias xvcfr='xvc file recheck'
alias xvcp='xvc pipeline'
alias xvcpr='xvc pipeline run'
alias xvcps='xvc pipeline step'
alias xvcpsn='xvc pipeline step new'
alias xvcpsd='xvc pipeline step dependency'
alias xvcpso='xvc pipeline step output'
alias xvcpi='xvc pipeline import'
alias xvcpe='xvc pipeline export'
alias xvcpl='xvc pipeline list'
alias xvcpn='xvc pipeline new'
alias xvcpu='xvc pipeline update'
alias xvcpd='xvc pipeline dag'
alias xvcs='xvc storage'
alias xvcsn='xvc storage new'
alias xvcsl='xvc storage list'
alias xvcsr='xvc storage remove'

🦊 We can use them when adding single-letter aliases.

🐇 There is a from_env method for Shell; will we support it?

🦊 I think we can support it. It’s much easier to use if we omit the shell.

🐲 We had aliases in xvc-core, but we need to access XvcCLI from completions. The completions module must be moved to xvc.

🦊 We added the clap_complete dependency to the xvc-pipeline and xvc-file crates, but I don’t think they are necessary. Let’s remove them now.

🐢 Now only the test-helper and xvc crates have the clap_complete dependency.

🐇 Are we ready to test?

🐢 Completions are working. 🎉

🐲 Now we need to update the docs and doc tests, I believe.

🐢 There are also tests to update.

🐇 Tests are running now. In the meantime, can we take a look at the blink configuration?

🐲 When we removed xvc aliases, we also removed pvc, xls, and other aliases. Maybe we can add these to the docs.

🐢 We can put them in the xvc completions reference for now.

🐢 It takes a while for rust-analyzer to finish analyzing the codebase.

🦊 Added aliases for xvc pipeline commands. Do you think we need to repeat root-level flags in xvc-pipeline?

🐢 No need to divert attention, I believe. Also, I still think there is an easier way to do that.

🐇 Okay. Do you think we should add easier subcommands to step? Like, step new becoming xvc p s n instead of xvc p s n? (Wait, that’s the same). I mean, more concise.

🐢 I think we can extend these even to the top-level, but shouldn’t make them visible. It will pollute the help text. We can add xvc fl for file list to avoid the space, but hide these from the help text.

🐲 That’s a good idea, but it will require including sub-crate level modules at the top level. We must test it first.

🐢 Let’s finish up the current changes and release them first.

🐇 By the way, we didn’t add two-letter abbreviations to the xvc storage new subcommands. Do you think we need them?

🐢 When we think about the frequency of these commands, no, I don’t think we need them. Users won’t add a new storage every day.

🐇 By that logic, we shouldn’t need an n for xvc storage new.

🐢 Actually, yeah. Maybe we should remove even s.

🦊 storage list may be useful.

🐢 s is a very common letter, though. We can have other uses for that letter.

🐇 Updating xvc p s dependency options, but it looks like making dependency options subcommands is a better way.

🦊 We didn’t do it because currently we can supply multiple options with a single command. If we go the subcommand route, we’ll have to write all dependencies one by one. Adding multiple subcommands with clap is a bit tricky.

🐢 Umm. Yeah, I see.

🐇 Maybe we can provide a separate command to add multiple dependencies. Like, xvc p s d add 'lines=myfile.csv::10-20; file=myimage.jpg; glob=dir/image-10*'

🐲 That looks like the start of a language. We need a parser for those strings. They will be freeform.

🦊 Another option is to keep the current options and add commands with the same names.

🐢 That will be confusing. The user will have both --param and param, and they will work differently.

🐇 We can have an add command that accepts the current options. Parsing will be done by clap just like now, but for a subcommand of dependency.

🐢 The full command will be something like xvc pipeline step --step-name preprocessing dependency add --params 'params.json::batch_size'

🐇 With shorter commands, it’s like xvc p s -s preprocessing d a --params 'params.json::batch_size', and this doesn’t look like ffmpeg monstrosities.

🐲 In any case, this is a backward-incompatible change. This should wait for v0.7 along with ECS changes.

🐢 ECS changes are not user-visible, but these are. Certainly needs a minor version update.

🐇 Then, okay, let’s keep the current ones for this version and think about updating them in a future version.

🐢 Okay. Let’s finish up and we’ll create a separate invisible subcommand to ask questions to the Xvc repo for completions.

🐇 We completed completions for the common command structure. Now, we need a way to show certain info after certain commands. A tab after xvc p r -p should show pipeline names, for example.

🐢 Umm, yeah. And these should be as quick as possible. They shouldn’t check Git or any other things.

🐇 I looked here and there, and the only working example I found is here: https://github.com/clap-rs/clap/blob/master/clap_complete/tests/testsuite/zsh.rs#L248 in clap_complete tests.

🐢 Let’s try to dive in. We can start by cloning the repo, I believe.