part: Check partition parsed partition type before setting it

[vojtechtrefny]

Summary by CodeRabbit

• Bug Fixes

  • Improved partition-type validation to detect and reject invalid numeric IDs and unknown partition types with clearer error messages, preventing invalid changes.

• Tests

  • Added tests covering additional invalid partition ID inputs to ensure validation catches malformed and out-of-range values.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

---

📝 Walkthrough

Walkthrough

Added stricter validation in set_part_type(): numeric parsing now verifies full-string consumption and rejects unknown MBR partition types; resources are released on validation failure. Also expanded tests with three invalid partition-id cases to assert error handling.

Changes

Cohort / File(s)	Summary
Partition Type Validation src/plugins/part.c	Added endptr-based numeric parsing check (uses g_ascii_strtoull and requires endptr at string terminator). After parsing, added fdisk_parttype_is_unknown() check to reject unknown MBR types; on failure unreferences ptype and pa and returns FALSE with an INVAL error.
Test Coverage Expansion tests/part_test.py	Added three invalid partition-id test cases ("83;id", "0xfff", "999") to PartSetIdCase to ensure invalid strings raise GLib.GError and are rejected.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 66.67% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately reflects the main change: adding validation to check the parsed partition type before setting it in the partition module.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 3

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@src/plugins/part.c`:
- Around line 1775-1781: The early-return path after calling
fdisk_parttype_is_unknown(ptype) leaks the parsed parttype; before returning
FALSE you must unref the allocated ptype by calling fdisk_unref_parttype(ptype)
(similar to the other early-return paths), so update the block that currently
calls fdisk_unref_partition(pa) and returns to also call
fdisk_unref_parttype(ptype) to free the resource while still setting the error
using type_str/BD_PART_ERROR_INVAL.
- Around line 1775-1781: The fdisk_parttype_is_unknown check is being applied
unconditionally and thus rejects structurally-valid custom GPT GUIDs; restrict
this validation to MSDOS/MBR only by moving or wrapping the
fdisk_parttype_is_unknown call (the block that sets BD_PART_ERROR_INVAL and
returns FALSE) so it only runs when table_type == BD_PART_TABLE_MSDOS (the same
scope where fdisk_label_parse_parttype is used for MBR); ensure
bd_part_set_part_type continues to accept unknown GPT parttypes parsed by
fdisk_label_parse_parttype and does not trigger the invalid-type error for GPT.

In `@tests/part_test.py`:
- Around line 1315-1318: The test using
BlockDev.part_set_part_id(self.loop_devs[0], ps.path, "83;id") is unreliable
because "83;id" can be parsed as decimal 83 (0x53) and may be present in
libfdisk’s MBR table; change the input to a value that cannot be interpreted as
a single-byte type such as "0x83;id" (or otherwise replace with an out-of-range
token like "0xfff") so fdisk_label_parse_parttype() will fail consistently and
the with self.assertRaises(GLib.GError): assertion around
BlockDev.part_set_part_id will reliably trigger; update or add a short comment
next to the test explaining the rationale.

[tbzatek]

Converting type_str to int back and forth is kinda nasty, but would require deeper refactoring. Good for now though.