atlass 1.13.0 → 2.0.0

package/README.md

@@ -2,7 +2,7 @@
 
 CLI for Atlassian Cloud. Views Jira issues in the terminal, copies Jira issues
 and Confluence pages to Markdown, pushes edits back, searches both, and shows
-Bitbucket pipeline runs.
+Bitbucket pipeline runs and pull requests.
 
 ## Install
 
@@ -31,8 +31,8 @@ Site and email go in `~/.config/atlass/config.json`. The token goes in the OS
 keyring. One account at a time.
 
 Bitbucket needs its own Bitbucket-scoped token with pipeline, pull request,
-account, and workspace read scopes, since Atlassian tokens are scoped per
-product:
+repository, account, and workspace read scopes, since Atlassian tokens are
+scoped per product:
 
 ```bash
 atlass bitbucket login    # prompts for workspace, default repo, and token
@@ -146,17 +146,29 @@ For both, `--json` and `--copy` work as for search.
 
 ```bash
 atlass jira search "safari login"
-atlass jira search --project PROJ --assignee me --status "In Progress"
-atlass jira search --jql "project = PROJ AND labels = regression"
-atlass confluence search "onboarding" --space DOCS
-atlass confluence search --cql "label = runbook ORDER BY created DESC"
+atlass jira search --project PROJ --assignee me --open
+atlass jira search --status "In Progress" --status "In Review" --updated 2w
+atlass confluence search "onboarding" --space DOCS --label runbook
 ```
 
-Filters combine with AND and sort by most recently updated. Rows use the same
-columns as `jira list` and `confluence list`. `--jql` and `--cql` replace the
-friendly filters. `--limit` defaults to 25, max 100. `--json` prints machine
-output. Add `--copy` to pick results interactively and copy each one, with
-`--out` as the target directory.
+Jira filters on `--project`, `--assignee`, `--reporter`, `--status`, `--type`,
+`--label`, `--updated` and `--open`; Confluence on `--space`, `--label`,
+`--updated` and `--starred`. Repeating a flag matches any of its values;
+different flags must all match. `--assignee` and `--reporter` take `me`, an
+account id, or a name to look up. `--updated` takes `7d`, `2w`, `3m` or
+`YYYY-MM-DD`. Results sort by most recently updated and use the same columns
+as `jira list` and `confluence list`.
+
+For anything the filters do not cover, write the query yourself:
+
+```bash
+atlass jira jql "project = PROJ AND labels = regression ORDER BY created DESC"
+atlass confluence cql "label = runbook AND creator = currentUser()"
+```
+
+`--limit` defaults to 25, max 100. `--json` prints machine output. Add `--copy`
+to pick results interactively and copy each one, with `--out` as the target
+directory. All four work on both search commands.
 
 Discovery aids for the filters:
 
@@ -191,6 +203,8 @@ atlass bitbucket prs --all                    # every state
 atlass bitbucket prs --reviewer me            # waiting on your review
 atlass bitbucket prs --author me --repo acme/web
 atlass bitbucket prs --query 'destination.branch.name = "main"'
+atlass bitbucket pr 842                       # one pull request in full
+atlass bitbucket pr https://bitbucket.org/acme/web/pull-requests/842
 ```
 
 `prs` shows `#ID  State  Age  Title`, most recently updated first, open only
@@ -201,6 +215,18 @@ unless `--state` or `--all` says otherwise. Drafts show as `DRAFT` in place of
 involving that person. `--query` replaces them and cannot be used alongside
 them, and it cannot mention `state`, which `--state` owns.
 
+`pr` shows one pull request: state, branches, approvals, the description, each
+reviewer and where they stand, the changed files with line counts, and the last
+five comment threads, inline ones anchored to `file:line`. A thread prints as its
+opening comment with every reply indented beneath it, a resolved one closes with
+who resolved it, and `@` mentions read as names. `--all-comments` shows the rest,
+`--no-pager` prints directly, and `--json` prints the whole thing.
+
+A pull request URL carries its own workspace and repo, so `--repo` is not needed
+alongside one. Long pull requests stop at 50 files and 200 comments. The file
+list needs the repository read scope; without it that section says so and the
+rest still prints.
+
 ## Development
 
 ```bash

package/dist/bitbucket.mjs

@@ -1,5 +1,5 @@
 #!/usr/bin/env node
-import { n as buildStandalone, r as fail } from "./build-BblaPBOA.mjs";
+import { n as buildStandalone, r as fail } from "./build-DMLlCHdF.mjs";
 //#region src/bitbucket.ts
 buildStandalone("bitbucket").parseAsync().catch(fail);
 //#endregion