(8a7fe6c) Read a flag by either spelling, on the prompt and in the fish rules, and bump to 0.33.8
`:mv --status ` offered the five statuses on the board and `:mv -s `
offered nothing: argument_at read the typed word with strip_prefix("--"),
so a word spelled with one hyphen was not a flag to it and fell through to
the positional arm, which counted a position that was not there.
long_name translates the letter through the command that holds the flag, so
the tables stay keyed on the long name and nothing here keeps a second list
of every option's letter. One letter, never a cluster: `mv -fs done` would
make the last letter the honest answer, but --title, --acceptance and a
comment body all take a leading hyphen on purpose, and a criterion typed as
`-fixes the offset` must not be answered with statuses.
The fish half had the same hole in its dynamic rules alone, which were
written as `-l <flag>`: `append -g <TAB>` offered nothing where
`append --tags <TAB>` offered the board's tags. fish_option reads the letter
off the definition per command, so an option given one tomorrow is offered
under it the day it is given. Clap's own generated rules already carried
both spellings, which is why a shell answered `mv -s ` and the board did
not.
a_short_flag_answers_what_its_long_spelling_answers walks every option with
both spellings on every command and nested command and pins that the two
answer the same, so this cannot come back one flag at a time.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Opus 5
parent
298e00eb33
commit
f1dc0efa6c
Generated
+1
-1
@@ -1563,7 +1563,7 @@ dependencies = [
|
||||
|
||||
[[package]]
|
||||
name = "worklane"
|
||||
version = "0.33.7"
|
||||
version = "0.33.8"
|
||||
dependencies = [
|
||||
"anyhow",
|
||||
"assert_cmd",
|
||||
|
||||
+1
-1
@@ -1,6 +1,6 @@
|
||||
[package]
|
||||
name = "worklane"
|
||||
version = "0.33.7"
|
||||
version = "0.33.8"
|
||||
edition = "2024"
|
||||
|
||||
[dependencies]
|
||||
|
||||
@@ -18,6 +18,8 @@ The generated half, the fish half, and what keeps the two honest.
|
||||
- **A `ValueEnum` is never listed twice, and `argument_at` is why the fish half's two hand-written value lists are the only ones.** clap emits the possible values of an *option*, so `mv --status` and `set --clear` need nothing here and the board reads them off the same definition with `Arg::get_possible_values`. The `ls` status and the `completions` shell are positionals, which the fish generator emits nothing for, so those two are written into the block from the enums — and the board reaches them through `get_positionals`, which needs no `Command::build` the way `get_index` would
|
||||
- **`RELATION_KIND_FLAGS` is the one position clap cannot answer for.** `relate --clear` takes a kind and then a target, two values of different kinds under one flag, so possible values on it would refuse every target. Both renderers therefore say so themselves: the fish rule lists `RelationKind::ALL` and the table names the pair
|
||||
- **`ls --ids` is the machine-readable listing**, and one of the three outputs in the tree meant to be parsed: the short id, a tab, the title, unstyled. Two fields because fish reads what follows a tab as the description and a script reads `cut -f1`. It includes done, which the decorated listing leaves out, because a completion has to be able to name a finished ticket, and it skips a missing status directory rather than refusing, because an error printed over a shell prompt is worse than no candidates. The filters still apply, and an unknown `--campaign` is still refused
|
||||
- **A flag is recognised by either spelling, `--status` and `-s` alike.** `long_name` translates the letter through the command that holds the flag, so every table above stays keyed on the long name and nothing here keeps a second list of the letters. The prompt read a flag by its two hyphens until then: `mv --status ` offered the five statuses and `mv -s ` offered nothing. **One letter, never a cluster** — `mv -fs done` would be the honest reading of the last letter, but `--title`, `--acceptance` and a comment body all take a leading hyphen on purpose, and a criterion typed as `-fixes the offset` must not be answered with statuses. `a_short_flag_answers_what_its_long_spelling_answers` walks every option with both spellings on every command and nested command, so an option given a letter tomorrow is covered the day it is given one
|
||||
- **The fish rules name the letter too, `-s g -l tags`.** The same hole on the other half: `append -g <TAB>` offered nothing where `append --tags <TAB>` offered the board's tags, because the dynamic rules were rendered with `-l` alone. `fish_option` reads the letter off the definition per command, so a rule for an option with none is what it always was. Clap's own generated half already emitted both spellings, which is why only the dynamic rules were short
|
||||
- **A flag's possible values are read off the command that *holds* the flag, nested or not.** `argument_at` resolves the group's child when the second word names one, so `campaign ls --state` reaches `ls`'s own arguments rather than `campaign`'s, which carry no such flag. Without that the prompt offered nothing where fish offered the three values, and the promise that a `ValueEnum` added anywhere is typeable on the board the day it is added held only for a top-level command
|
||||
- **`campaign ls --ids` is the same flag and the same two fields for the other object**, and it deliberately skips the member counting the decorated listing does: `members()` reads every ticket on the board, and the caller here is a shell filling in candidates on a keypress. It also prints nothing on an empty board, where the decorated listing prints a sentence, because a sentence offered as a candidate is worse than no candidates
|
||||
- **`worklane tags` is the third, and it is a command rather than a mode on `ls`.** Two reasons: `ls --tag <tag>` answers what carries a tag and cannot answer what tags exist, which is the question a grooming sweep starts from, and a `--tags` beside a `--tag` is one letter from a filter that means something else. Alphabetical and case sensitive, matching the filter, so `bug` and `Bug` list separately and land next to each other, which is where a near-duplicate is spotted. Every status, done included: a tag that has fallen out of use is still part of what the vocabulary has been. `assets/PLAN.md` used to tell the agent to count tags with a `show` per ticket, and now points here
|
||||
|
||||
+148
-8
@@ -111,6 +111,36 @@ pub const TAG_NESTED_COMMANDS: [(&str, &str); 1] = [("tags", "set")];
|
||||
/// every target.
|
||||
pub const RELATION_KIND_FLAGS: [(&str, &str); 1] = [("relate", "clear")];
|
||||
|
||||
/// The long name of the flag a typed word spells, or nothing where it is not a flag.
|
||||
///
|
||||
/// `--status` is its own name and `-s` is clap's to translate, off the command that holds
|
||||
/// the flag: a table of the shorts kept here would be a second list of every option the
|
||||
/// tool has, and it would go wrong the first time one was given a letter.
|
||||
///
|
||||
/// **One letter, never a cluster.** `mv -fs done` is force and status and it is `status`
|
||||
/// that takes the word after it, so the last letter of a cluster would be the honest
|
||||
/// answer — but `--title`, `--description`, `--acceptance`, `--plan` and a comment body all
|
||||
/// take a leading hyphen on purpose, and reading `-fixes the offset` as a cluster would
|
||||
/// offer statuses inside a criterion. A clustered short is rare on a board where the line
|
||||
/// is typed a word at a time; a hyphen-leading prose value is a documented feature.
|
||||
fn long_name(arg: &str, holder: Option<&clap::Command>) -> Option<String> {
|
||||
if let Some(long) = arg.strip_prefix("--") {
|
||||
return (!long.is_empty()).then(|| long.to_string());
|
||||
}
|
||||
|
||||
let mut letters = arg.strip_prefix('-')?.chars();
|
||||
let short = match (letters.next(), letters.next()) {
|
||||
(Some(letter), None) => letter,
|
||||
_ => return None,
|
||||
};
|
||||
|
||||
holder?
|
||||
.get_arguments()
|
||||
.find(|option| option.get_short() == Some(short))
|
||||
.and_then(|option| option.get_long())
|
||||
.map(str::to_string)
|
||||
}
|
||||
|
||||
/// What an argument of a typed line names, where worklane can answer what goes there.
|
||||
///
|
||||
/// The board's command prompt completes against this: the tables above say *what* the
|
||||
@@ -178,8 +208,10 @@ pub fn argument_at(args: &[String], index: usize) -> Option<Argument> {
|
||||
.or(group);
|
||||
|
||||
// A flag's value is the argument after it, which is where the flag sits whether the
|
||||
// value is half typed or has not been started.
|
||||
if let Some(flag) = args.get(index - 1).and_then(|arg| arg.strip_prefix("--")) {
|
||||
// value is half typed or has not been started. Spelled long or short: `-s` is the same
|
||||
// flag as `--status` and the row has to answer both, which is what `long_name` is for.
|
||||
if let Some(flag) = args.get(index - 1).and_then(|arg| long_name(arg, holder)) {
|
||||
let flag = flag.as_str();
|
||||
if TICKET_FLAGS.contains(&(verb, flag)) {
|
||||
return Some(Argument::Ticket);
|
||||
}
|
||||
@@ -281,8 +313,9 @@ fn fish_dynamic() -> String {
|
||||
let ticket_flags = TICKET_FLAGS
|
||||
.iter()
|
||||
.map(|(command, flag)| {
|
||||
let option = fish_option(command, flag);
|
||||
format!(
|
||||
"complete -c worklane -n '__worklane_command_is {command}' -l {flag} \\\n \
|
||||
"complete -c worklane -n '__worklane_command_is {command}' {option} \\\n \
|
||||
-x -a '(worklane ls --ids)'"
|
||||
)
|
||||
})
|
||||
@@ -299,12 +332,13 @@ fn fish_dynamic() -> String {
|
||||
.map(|(parent, child)| format!("__worklane_nested_arg_of {parent} {child}"))
|
||||
.collect::<Vec<_>>()
|
||||
.join("; or ");
|
||||
let campaign_flags = flag_list(&CAMPAIGN_FLAGS);
|
||||
let campaign_flags = flag_list("ls", &CAMPAIGN_FLAGS);
|
||||
let tag_flags = TAG_FLAGS
|
||||
.iter()
|
||||
.map(|(command, flag)| {
|
||||
let option = fish_option(command, flag);
|
||||
format!(
|
||||
"complete -c worklane -n '__worklane_command_is {command}' -l {flag} \\\n \
|
||||
"complete -c worklane -n '__worklane_command_is {command}' {option} \\\n \
|
||||
-x -a '(worklane tags | cut -f1,3)'"
|
||||
)
|
||||
})
|
||||
@@ -442,15 +476,40 @@ complete -c worklane -n '__worklane_first_arg_of install-skill' -a '{skills}' -d
|
||||
)
|
||||
}
|
||||
|
||||
/// Option names as fish wants them on a `complete` line, `-l one -l two`.
|
||||
fn flag_list(flags: &[&str]) -> String {
|
||||
/// Option names as fish wants them on a `complete` line, `-s o -l one -l two`, for the
|
||||
/// options of one command.
|
||||
fn flag_list(command: &str, flags: &[&str]) -> String {
|
||||
flags
|
||||
.iter()
|
||||
.map(|flag| format!("-l {flag}"))
|
||||
.map(|flag| fish_option(command, flag))
|
||||
.collect::<Vec<_>>()
|
||||
.join(" ")
|
||||
}
|
||||
|
||||
/// One option as fish names it: its long name, and its letter where clap gave it one.
|
||||
///
|
||||
/// `-s g -l tags` rather than `-l tags` alone, or `append -g <TAB>` offers nothing where
|
||||
/// `append --tags <TAB>` offers the board's tags — the same hole the board's own prompt had
|
||||
/// until `long_name`, and the reason one table answers both halves is that neither may
|
||||
/// answer a spelling the other does not. The letter is read off the definition and never
|
||||
/// written here, so an option given one tomorrow is offered under it the day it is given.
|
||||
fn fish_option(command: &str, flag: &str) -> String {
|
||||
let cli = crate::command::Cli::command();
|
||||
let short = cli
|
||||
.find_subcommand(command)
|
||||
.and_then(|command| {
|
||||
command
|
||||
.get_arguments()
|
||||
.find(|option| option.get_long() == Some(flag))
|
||||
})
|
||||
.and_then(|option| option.get_short());
|
||||
|
||||
match short {
|
||||
Some(letter) => format!("-s {letter} -l {flag}"),
|
||||
None => format!("-l {flag}"),
|
||||
}
|
||||
}
|
||||
|
||||
/// The names a `ValueEnum` accepts, for a fish `-a` list.
|
||||
///
|
||||
/// Read off the enum rather than written out, so a status or a shell added to either one
|
||||
@@ -774,6 +833,87 @@ mod tests {
|
||||
assert!(shells.contains(&"fish".to_string()), "{shells:?}");
|
||||
}
|
||||
|
||||
/// The fish half names an option by its letter too, wherever clap gave it one, so a
|
||||
/// dynamic rule answers both spellings the way the prompt does.
|
||||
#[test]
|
||||
fn a_fish_rule_names_the_option_by_its_letter_as_well() {
|
||||
let script = fish_dynamic();
|
||||
|
||||
// `append --tags` is the one flag in these tables with a letter, and it is the
|
||||
// whole reason the rule is rendered rather than written: `-g` offered nothing.
|
||||
assert!(
|
||||
script.contains("-s g -l tags"),
|
||||
"the tags rule does not name its letter:\n{script}"
|
||||
);
|
||||
|
||||
// And one without stays as it was, rather than gaining an empty `-s`.
|
||||
assert!(script.contains("-l campaign"), "{script}");
|
||||
assert!(!script.contains("-s "), "an option with no letter got one:\n{script}");
|
||||
|
||||
// Read off the definition, never written here.
|
||||
assert_eq!(fish_option("append", "tags"), "-s g -l tags");
|
||||
assert_eq!(fish_option("ls", "tag"), "-l tag");
|
||||
assert_eq!(fish_option("ls", "nosuchflag"), "-l nosuchflag");
|
||||
}
|
||||
|
||||
/// A flag spelled short answers exactly what the same flag spelled long answers, on
|
||||
/// every command and every option that has both.
|
||||
///
|
||||
/// Walked off the clap definition rather than listed, so an option given a letter
|
||||
/// tomorrow is covered the day it is given one. `-s` was the report: `mv --status `
|
||||
/// offered the five statuses and `mv -s ` offered nothing, because the row read a flag
|
||||
/// by its two hyphens. A boolean's two spellings agree as well, both answering nothing,
|
||||
/// which is the same statement about a value that is not there.
|
||||
#[test]
|
||||
fn a_short_flag_answers_what_its_long_spelling_answers() {
|
||||
let cli = crate::command::Cli::command();
|
||||
let mut seen = 0;
|
||||
|
||||
for command in cli.get_subcommands() {
|
||||
let children: Vec<(String, &clap::Command)> = command
|
||||
.get_subcommands()
|
||||
.map(|child| (format!("{} {}", command.get_name(), child.get_name()), child))
|
||||
.collect();
|
||||
|
||||
for (line, holder) in
|
||||
std::iter::once((command.get_name().to_string(), command)).chain(children)
|
||||
{
|
||||
for option in holder.get_arguments() {
|
||||
let (Some(short), Some(long)) = (option.get_short(), option.get_long()) else {
|
||||
continue;
|
||||
};
|
||||
|
||||
seen += 1;
|
||||
assert_eq!(
|
||||
at_cursor(&format!("{line} -{short} ")),
|
||||
at_cursor(&format!("{line} --{long} ")),
|
||||
"`-{short}` and `--{long}` answer differently on `{line}`"
|
||||
);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
assert!(seen > 20, "only {seen} options had both spellings, which is too few");
|
||||
}
|
||||
|
||||
/// The two the report named, spelled out, so the walk above cannot pass by finding
|
||||
/// nothing on either spelling.
|
||||
#[test]
|
||||
fn a_short_flag_reaches_the_values_and_the_board() {
|
||||
let listed: Vec<String> = Status::ALL.iter().map(|s| s.as_str().to_string()).collect();
|
||||
assert_eq!(values("mv -s "), Some(listed), "`-s` offers no statuses");
|
||||
assert_eq!(values("mv -s bac"), values("mv --status bac"), "half typed");
|
||||
|
||||
// A tag, which comes off the board rather than out of the enum.
|
||||
assert_eq!(at_cursor("append -g "), Some(Argument::Tag { declared: false }));
|
||||
|
||||
// And a cluster is deliberately not read, because `--title`, `--acceptance` and a
|
||||
// comment body all take a leading hyphen: a value spelled `-fixes the offset` must
|
||||
// not be read as `-s` and answered with statuses.
|
||||
assert_eq!(at_cursor("mv -fs "), None);
|
||||
assert_eq!(at_cursor("comment 1b78a24 -fs "), None);
|
||||
}
|
||||
|
||||
/// The flag on a nested command is answered the same way, from the enum: nothing here
|
||||
/// knows what a campaign's states are, and the values reach the prompt and fish alike
|
||||
/// the day a fourth one is added to `CampaignState`.
|
||||
|
||||
Reference in New Issue
Block a user