swev-id: astropy__astropy-14365 — ascii.qdp: accept case-insensitive QDP commands; add tests

[rowan-stein]

Updated the ascii.qdp reader to recognize QDP error command directives (READ SERR/READ TERR) case-insensitively using an inline regex flag scoped only to the commands. Added tests covering uppercase, lowercase, and mixed-case inputs. No changes to data parsing or table splitting tokens. Fixes #26.

[noa-lucent]

Thanks for addressing Issue #26. The approach to scope case-insensitivity via an inline regex flag is sound and should limit impact to command detection. However, the PR currently has blocking issues:

• _command_re string in qdp.py is syntactically incomplete; the inline flag group appears unclosed.
• The new test file starts with a leading space and ends with non-Python content (PYTEST and shell commands), which will cause syntax errors.

Please fix these and consider the minor regex improvements suggested. Once addressed, I’m happy to re-review and approve.

[noa-lucent]

[major] The test file includes non-Python content at the end (lines like PYTEST and shell commands: git add -A, gh pr create, etc.). These will cause syntax errors and must be removed. The file should only contain test code.

[noa-lucent]

[minor] Nice coverage of uppercase/lowercase/mixed cases and with/without explicit names. To strengthen robustness, consider adding tests for:

• Tabs or multiple spaces between tokens (e.g., READ\tSERR\t3 or READ TERR 1)
• Trailing whitespace on directive lines (READ SERR 3 )
• A comment line following directives to ensure comment detection remains unaffected

These help confirm delimiter handling and that case-insensitive matching doesn’t interfere with other line-type detection.

[noa-lucent]

[major] The _command_re assignment appears syntactically incomplete. The inline flag group (?i:...) is missing its closing parenthesis, and the string literal lacks a closing quote. This will cause a syntax error and prevent parsing. Please fix to something like:

• Minimal fix: r"(?i:READ [TS]ERR(\s+[0-9]+)+)"
• More robust (anchor and allow trailing spaces): r"(?i)^READ [TS]ERR(?:\s+[0-9]+)+\s*$"

The inline flag is appropriately scoped to the command portion, so this change should only affect command matching and leave decimal/data/comment/new-table detection unchanged.

[noa-lucent]

[major] Unexpected indent before top-level imports (leading space at column 1). This will raise IndentationError. Please remove the leading space so imports start at column 0.

[noa-lucent]

[minor] Consider anchoring the command regex and permitting optional leading/trailing whitespace to avoid false positives/negatives, e.g. r"(?i)^\s*READ [TS]ERR(?:\s+[0-9]+)+\s*$". This prevents a data line that happens to contain the substring from matching, and tolerates trailing spaces which are common in QDP files.

[noa-lucent]

Thanks for the update. Verification on the latest commit (82c020b):

• qdp.py: _command_re now correctly uses an inline, scoped case-insensitive group: (?i:READ\s+[TS]ERR(?:\s+[0-9]+)+). No conflict markers present. Other patterns and line-type logic appear unchanged.
• Tests: Added coverage for uppercase/lowercase/mixed-case, tabs/multiple spaces, trailing whitespace, and a comment line after directives — assertions look correct. However, the test file currently has top-level indentation (one leading space) before def _make_qdp_text, the @pytest.mark.parametrize decorator, and def test_qdp_command_case_insensitive, which will raise an IndentationError. There are also trailing spaces on blank lines after imports.

Requested fixes:

1. Remove the leading space at the start of all top-level decorators and def statements so they begin at column 0.
2. Strip trailing whitespace on blank lines after the import block.

Once these are addressed, I’m ready to approve.

[noa-lucent]

[nit] The note below the _command_re repeats the case-insensitivity statement already conveyed in the preceding comment. Consider condensing/removing the redundant line to keep comments crisp.

[noa-lucent]

[nit] Minor style: the two blank lines after the imports contain trailing spaces. Please strip trailing whitespace to satisfy linters.

[noa-lucent]

[major] Several top-level statements are indented by one space (e.g., def _make_qdp_text(...), the @pytest.mark.parametrize(...) decorator, and the subsequent def test_qdp_command_case_insensitive(...)). Top-level defs/decorators must start at column 0; this will raise an IndentationError and break the test file. Please remove the leading spaces so these lines are flush-left.

[rowan-stein]

Closing to enforce single open PR per task (ID: 10). Please refer to PR #121 for the active work.