feat(test): parse XML context files and generate tree diagrams

[tfcollins]

Summary

Added support for parsing each XML context file under the test scope and producing a corresponding tree diagram representation. This improves visibility into XML structure and makes context validation easier during development and debugging.

Changes

• Implemented XML parsing flow for all context files located under test directories.
• Added tree diagram generation for each parsed XML context file.
• Standardized output structure so diagrams are consistent and easier to compare across files.
• Integrated the workflow into the existing test-oriented process so diagram generation can be run as part of validation.

Impact

• Improves readability of complex XML contexts by surfacing hierarchy in a concise visual format.
• Reduces manual effort when inspecting or troubleshooting context file structure.
• Helps catch structural inconsistencies earlier in test and review cycles.

Notes for Reviewers

• Focus on parser behavior across different XML context file shapes (nested nodes, attributes, mixed depth).
• Verify tree output is deterministic and stable for repeated runs.
• Confirm generated diagrams align with expected XML hierarchy for representative test fixtures.

Closes https://github.com/tfcollins/tbot/issues/26

[github-actions]

Test Results

1 742 tests  +2     484 ✅ +2   12m 19s ⏱️ +6s
    1 suites ±0   1 258 💤 ±0 
    1 files   ±0       0 ❌ ±0 

Results for commit eb34adc. ± Comparison against base commit cfff950.

This pull request removes 4 and adds 6 tests. Note that renamed tests count towards both.

test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=rx_hardwaregain_chan0-val=[np.float64(13.0), np.float64(26.5), np.float64(20.5), np.float64(10.0), np.float64(26.5), np.float64(11.5), np.float64(14.0), np.float64(13.0), np.float64(14.0), np.float64(29.0)]-depends={'gain_control_mode_chan1': 'spi', 'rx_ensm_mode_chan1': 'rf_enabled'}-classname=adi.adrv9002]
test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=rx_hardwaregain_chan0-val=[np.float64(15.0), np.float64(24.5), np.float64(16.5), np.float64(22.5), np.float64(5.0), np.float64(1.0), np.float64(24.0), np.float64(13.5), np.float64(2.5), np.float64(34.0)]-depends={'gain_control_mode_chan0': 'spi', 'rx_ensm_mode_chan0': 'rf_enabled'}-classname=adi.adrv9002]
test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=tx_hardwaregain_chan0-val=[np.float64(-11.45), np.float64(-2.95), np.float64(-28.2), np.float64(-36.8), np.float64(-5.4), np.float64(-17.65), np.float64(-29.35), np.float64(-0.95), np.float64(-7.8), np.float64(-0.95)]-depends={'atten_control_mode_chan0': 'spi'}-classname=adi.adrv9002]
test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=tx_hardwaregain_chan1-val=[np.float64(-13.95), np.float64(-18.95), np.float64(-20.7), np.float64(-14.95), np.float64(-38.7), np.float64(-21.2), np.float64(-29.6), np.float64(-19.2), np.float64(-6.2), np.float64(-14.15)]-depends={'atten_control_mode_chan1': 'spi'}-classname=adi.adrv9002]

test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=rx_hardwaregain_chan0-val=[np.float64(10.5), np.float64(19.0), np.float64(13.0), np.float64(23.0), np.float64(12.5), np.float64(13.0), np.float64(13.5), np.float64(15.5), np.float64(29.0), np.float64(28.0)]-depends={'gain_control_mode_chan0': 'spi', 'rx_ensm_mode_chan0': 'rf_enabled'}-classname=adi.adrv9002]
test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=rx_hardwaregain_chan0-val=[np.float64(11.5), np.float64(15.5), np.float64(27.0), np.float64(33.0), np.float64(14.5), np.float64(21.5), np.float64(0.0), np.float64(32.5), np.float64(9.0), np.float64(15.0)]-depends={'gain_control_mode_chan1': 'spi', 'rx_ensm_mode_chan1': 'rf_enabled'}-classname=adi.adrv9002]
test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=tx_hardwaregain_chan0-val=[np.float64(-39.75), np.float64(-5.4), np.float64(-32.0), np.float64(-25.35), np.float64(-4.8), np.float64(-36.4), np.float64(-3.7), np.float64(-2.05), np.float64(-17.0), np.float64(-7.95)]-depends={'atten_control_mode_chan0': 'spi'}-classname=adi.adrv9002]
test.test_adrv9002_p ‑ test_adrv9002_hardware_gain[attr=tx_hardwaregain_chan1-val=[np.float64(-24.0), np.float64(-21.2), np.float64(-30.35), np.float64(-14.25), np.float64(-14.95), np.float64(-17.35), np.float64(-34.9), np.float64(-20.45), np.float64(-0.85), np.float64(-37.55)]-depends={'atten_control_mode_chan1': 'spi'}-classname=adi.adrv9002]
test.test_emu_xml_docgen ‑ test_generation_is_deterministic_and_cleans_stale_files
test.test_emu_xml_docgen ‑ test_tree_from_real_xml_contains_expected_structure

♻️ This comment has been updated with latest results.

[github-actions]

Generated documentation for this PR is available at Link

[gastmaier]

Sounds good, but why not have a small directive that calls doc/gen_emu_xml_trees.py ?
Like this (or simpler, just invoke the generation directly if cache is not cared):
https://github.com/analogdevicesinc/doctools/blob/main/adi_doctools/directive/hdl.py#L405-L446

instead of committing all svgs

Another option is having, instead of xml -> svg, have xml -> mermaid, and let the very well tested flowchart library handle the details