test: complete goto_reply fixture and assertion (#11)#16
Conversation
Was hardcoded to 0.12.2; glob-match any Cellar/neovim/*/share tree so the tests keep running after nvim upgrades. Falls back to /tmp/plenary.nvim for CI where the Cellar path does not exist.
- Extract collect_quoted_sig_ranges(lines) as a pure function that collects all '> --' ranges at any quoting depth (^>[ >]*%-%-). Testable without a real buffer. - kill_quoted_sig iterates collected ranges bottom-to-top so earlier line indices remain valid after each nvim_buf_set_lines call. - Broadened separator pattern from '^> ?%-%-' to '^>[ >]*%-%-' to handle doubly-quoted (and deeper) sig separators in long threads. - Add two new tests: multi-block buffer test (fails old code, passes new) and a pure-function range-collection unit test. - Add fixture draft_multi_quoted_sig.txt for reference.
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request refactors contact resolution, querying, and picking to run asynchronously using vim.system, extracting the resolver logic into a dedicated resolver.lua module. It also updates the calendar integration to use event-specific dates rather than the current date, improves signature stripping to handle multiple quoted signatures in long threads, and adds corresponding unit tests. The review feedback highlights two issues: a potential runtime error in contacts.lua if JSON decoding returns a non-table value, and a pattern matching bug in contacts_picker.lua that silently filters out contacts without a display name.
Important
The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.
| local ok, data = pcall(vim.json.decode, output) | ||
| if not ok or not data then return {} end |
There was a problem hiding this comment.
If output is valid JSON but parses to a non-table value (such as a number, string, or boolean), data will not be a table. Calling ipairs(data) on line 87 will then throw a runtime error.
Using type(data) ~= 'table' instead of not data ensures robust handling of any unexpected JSON formats.
local ok, data = pcall(vim.json.decode, output)
if not ok or type(data) ~= 'table' then return {} end
| local entries, seen = {}, {} | ||
| for _, line in ipairs(vim.split(stdout or '', '\n')) do | ||
| if line ~= '' then | ||
| local email, name = line:match('^([^\t]+)\t([^\t]+)') |
There was a problem hiding this comment.
The pattern ^([^\\t]+)\\t([^\\t]+) requires at least one non-tab character for the name field. If a contact has no display name (e.g., the line is email\\t or email\\t\\t), the match will fail entirely and return nil, causing these contacts to be silently omitted from the Telescope picker.
Changing the pattern to ^([^\\t]+)\\t([^\\t]*) allows matching empty name fields, returning "" for the name and correctly preserving the contact.
local email, name = line:match('^([^\\t]+)\t([^\\t]*)')
1 participant
Closes #11
Add
tests/mail/fixtures/draft_with_quotes.txtwith a realistic reply draft containing>-quoted lines, and rewrite the placeholder test to load it, callnav.goto_reply(), and assert the cursor lands on a line matching^>.All 82 tests pass.