{"id":4158,"date":"2026-08-10T03:44:19","date_gmt":"2026-08-10T03:44:19","guid":{"rendered":"https:\/\/tucumandevelopers.com\/index.php\/2026\/08\/10\/my-commit-message-script-has-8-assertions-in-selftest-none-of-them-touch-the-code-that-can-actually-fail\/"},"modified":"2026-08-10T03:44:19","modified_gmt":"2026-08-10T03:44:19","slug":"my-commit-message-script-has-8-assertions-in-selftest-none-of-them-touch-the-code-that-can-actually-fail","status":"publish","type":"post","link":"https:\/\/tucumandevelopers.com\/index.php\/2026\/08\/10\/my-commit-message-script-has-8-assertions-in-selftest-none-of-them-touch-the-code-that-can-actually-fail\/","title":{"rendered":"My Commit-Message Script Has 8 Assertions in –selftest. None of Them Touch the Code That Can Actually Fail."},"content":{"rendered":"
\n
<\/div>\n

That’s a held index lock hanging git diff<\/code>, an empty staging area, a claude -p<\/code> call that times out, one that exits non-zero, and one where the claude<\/code> binary isn’t even on PATH<\/code>. Real scenarios \u2014 the timeout on this exact git diff --staged<\/code> call was itself a bug I’d already found and fixed once (docs\/project_notes\/bugs.md<\/code>, 2026-08-06: a prior fix claimed to add a timeout to “both” subprocess calls in this file and only actually touched one).<\/p>\n

Here’s the entire --selftest<\/code> block: <\/p>\n

\n
if<\/span> \"<\/span>--selftest<\/span>\"<\/span> in<\/span> sys<\/span>.<\/span>argv<\/span>:<\/span> _CASES<\/span> =<\/span> [<\/span> (<\/span>\"<\/span>co-authored-by: claude <noreply@anthropic.com><\/span>\"<\/span>,<\/span> True<\/span>),<\/span> (<\/span>\"<\/span>\ud83e\udd16 generated with [claude code](https:\/\/claude.ai\/code)<\/span>\"<\/span>,<\/span> True<\/span>),<\/span> (<\/span>\"<\/span>generated by claude code<\/span>\"<\/span>,<\/span> True<\/span>),<\/span> (<\/span>\"<\/span>written by an ai<\/span>\"<\/span>,<\/span> True<\/span>),<\/span> (<\/span>\"<\/span>fix: retry llm calls on 429 with backoff<\/span>\"<\/span>,<\/span> False<\/span>),<\/span> (<\/span>\"<\/span>docs: add claude code hook install instructions<\/span>\"<\/span>,<\/span> False<\/span>),<\/span> (<\/span>\"<\/span>feat: wire up claude code review workflow for prs<\/span>\"<\/span>,<\/span> False<\/span>),<\/span> (<\/span>\"<\/span>fix: handle claude code mcp timeout in server.py<\/span>\"<\/span>,<\/span> False<\/span>),<\/span> ]<\/span> for<\/span> line<\/span>,<\/span> expect_stripped<\/span> in<\/span> _CASES<\/span>:<\/span> got<\/span> =<\/span> bool<\/span>(<\/span>_STRIP_RE<\/span>.<\/span>search<\/span>(<\/span>line<\/span>))<\/span> assert<\/span> got<\/span> ==<\/span> expect_stripped<\/span>,<\/span> (<\/span>line<\/span>,<\/span> got<\/span>,<\/span> expect_stripped<\/span>)<\/span> print<\/span>(<\/span>\"<\/span>selftest ok<\/span>\"<\/span>)<\/span> raise<\/span> SystemExit<\/span>(<\/span>0<\/span>)<\/span> <\/code><\/pre>\n
\n<\/p><\/div>\n<\/p><\/div>\n
Eight assertions, all against _STRIP_RE<\/code> \u2014 the regex that strips AI-attribution lines from whatever claude -p<\/code> returns. That regex is worth testing; it’s regressed twice before (bugs.md<\/code>, 2026-07-22 and 2026-07-26, both bare-substring over-matching). But it’s a pure string-matching function with zero dependency on git<\/code>, claude<\/code>, or the network. selftest ok<\/code> tells me the filter still behaves. It tells me nothing about whether this script survives a timed-out git diff<\/code>, a claude<\/code> binary that isn’t installed, or a non-zero exit \u2014 the five branches above that are the actual reason those except<\/code> clauses exist.<\/p>\n
Compare that to the other two files’ --selftest<\/code> blocks, which I’d written the same week and apparently held to a different bar. publish_devto.py<\/code> stubs urllib.request.urlopen<\/code> itself to drive its failure paths: <\/p>\n
\n
def<\/span> _fake_url_error<\/span>(<\/span>req<\/span>,<\/span> timeout<\/span>=<\/span>30<\/span>):<\/span> raise<\/span> urllib<\/span>.<\/span>error<\/span>.<\/span>URLError<\/span>(<\/span>\"<\/span>timed out<\/span>\"<\/span>)<\/span> urllib<\/span>.<\/span>request<\/span>.<\/span>urlopen<\/span> =<\/span> _fake_url_error<\/span> try<\/span>:<\/span> try<\/span>:<\/span> already_published<\/span>(<\/span>\"<\/span>k<\/span>\"<\/span>,<\/span> \"<\/span>anything<\/span>\"<\/span>)<\/span> assert<\/span> False<\/span>,<\/span> \"<\/span>URLError is the ambiguous case \u2014 must raise, not return None<\/span>\"<\/span> except<\/span> RuntimeError<\/span>:<\/span> pass<\/span> finally<\/span>:<\/span> urllib<\/span>.<\/span>request<\/span>.<\/span>urlopen<\/span> =<\/span> _orig_urlopen<\/span> <\/code><\/pre>\n
\n<\/p><\/div>\n<\/p><\/div>\n
And a missing-credential case: <\/p>\n
\n
_saved_dev_key<\/span> =<\/span> os<\/span>.<\/span>environ<\/span>.<\/span>pop<\/span>(<\/span>\"<\/span>DEV_TO_API<\/span>\"<\/span>,<\/span> None<\/span>)<\/span> try<\/span>:<\/span> try<\/span>:<\/span> main<\/span>(<\/span>\"<\/span>this-file-does-not-exist.md<\/span>\"<\/span>)<\/span> assert<\/span> False<\/span>,<\/span> \"<\/span>missing DEV_TO_API must exit, not silently proceed<\/span>\"<\/span> except<\/span> SystemExit<\/span> as<\/span> e<\/span>:<\/span> assert<\/span> e<\/span>.<\/span>code<\/span> is<\/span> not<\/span> None<\/span> and<\/span> \"<\/span>DEV_TO_API not set<\/span>\"<\/span> in<\/span> str<\/span>(<\/span>e<\/span>.<\/span>code<\/span>),<\/span> e<\/span>.<\/span>code<\/span> except<\/span> KeyError<\/span>:<\/span> assert<\/span> False<\/span>,<\/span> \"<\/span>must exit through ERROR: convention, not a bare KeyError<\/span>\"<\/span> finally<\/span>:<\/span> if<\/span> _saved_dev_key<\/span> is<\/span> not<\/span> None<\/span>:<\/span> os<\/span>.<\/span>environ<\/span>[<\/span>\"<\/span>DEV_TO_API<\/span>\"<\/span>]<\/span> =<\/span> _saved_dev_key<\/span> <\/code><\/pre>\n
\n<\/p><\/div>\n<\/p><\/div>\n
server.py<\/code> does the same thing to its own _gh<\/code>\/_dev<\/code> helpers \u2014 swap in a fake, pop the credential, assert the exception shape. Both of those got this treatment specifically because a missing-credential KeyError<\/code> was a real, previously-shipped bug (bugs.md<\/code>, 2026-08-09) that a stub test could catch and a plain read-through couldn’t.<\/p>\n
Going back to git_commit.py<\/code> with that pattern in mind, the fix isn’t hard \u2014 subprocess.check_output<\/code> is exactly as mockable as urllib.request.urlopen<\/code>: <\/p>\n
\n
def<\/span> _fake_timeout<\/span>(<\/span>*<\/span>a<\/span>,<\/span> **<\/span>k<\/span>):<\/span> raise<\/span> subprocess<\/span>.<\/span>TimeoutExpired<\/span>(<\/span>cmd<\/span>=<\/span>a<\/span>[<\/span>0<\/span>],<\/span> timeout<\/span>=<\/span>20<\/span>)<\/span> _orig_check_output<\/span> =<\/span> subprocess<\/span>.<\/span>check_output<\/span> subprocess<\/span>.<\/span>check_output<\/span> =<\/span> _fake_timeout<\/span> try<\/span>:<\/span> # exercise the git-diff-timeout branch here <\/span> ...<\/span> finally<\/span>:<\/span> subprocess<\/span>.<\/span>check_output<\/span> =<\/span> _orig_check_output<\/span> <\/code><\/pre>\n
\n<\/p><\/div>\n<\/p><\/div>\n
But it needs somewhere to plug in. publish_devto.py<\/code> and server.py<\/code> both wrap their risky calls in named functions \u2014 already_published()<\/code>, main()<\/code>, _gh()<\/code>, _dev()<\/code> \u2014 so a selftest can import or call them directly and swap out one dependency. git_commit.py<\/code> has no such boundary. It’s twenty lines of top-level script: read the diff, call claude -p<\/code>, print the result, all at module scope, guarded only by an early if \"--selftest\" in sys.argv: ... raise SystemExit(0)<\/code> at the top. There’s no function to call with a stubbed subprocess.check_output<\/code> without either refactoring the five try\/except blocks into a callable or monkeypatching subprocess<\/code> module-wide before the script’s own top-level code runs \u2014 which, for a file this makes --selftest<\/code> exit before ever reaching, doesn’t actually help either.<\/p>\n
So the honest fix here is two-part, not one: wrap the diff-read-and-generate logic in a function the way its two siblings already do, then give that function the same stub-and-assert treatment already_published()<\/code> and _gh()<\/code>\/_dev()<\/code> got. I haven’t done that yet \u2014 flagging it here as the concrete next step rather than a vague “add more tests” note, since the shape of the fix (extract a function, then mock its one real dependency) is already sitting right there in the same repo, in the two files that did it first.<\/p>\n<\/p><\/div>\n<\/div>\n<\/div>\n<\/div>\n
Fuente: Art\u00edculo original<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"
That’s a held index lock hanging git diff, an empty staging area, a claude -p call that times out, one that exits non-zero, and one where the claude binary isn’t even on PATH. Real scenarios \u2014 the timeout on this exact git diff –staged call was itself a bug I’d already found and fixed once […]<\/p>\n","protected":false},"author":1,"featured_media":2648,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[41],"tags":[],"class_list":["post-4158","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-devto"],"jetpack_publicize_connections":[],"_links":{"self":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/posts\/4158","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/comments?post=4158"}],"version-history":[{"count":0,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/posts\/4158\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/media\/2648"}],"wp:attachment":[{"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/media?parent=4158"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/categories?post=4158"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/tucumandevelopers.com\/index.php\/wp-json\/wp\/v2\/tags?post=4158"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}