[PG1204]

What changes were proposed in this PR?

If the HuggingFace operator's Input Image Column (or Input Audio Column) named a column that doesn't exist in the input table, use_image_column / use_audio_column silently became false and the user got the misleading "No image source. Set an Input Image Column or upload an image.", even though they had set one, just misspelled it.

This distinguishes the two cases: at that point there's no upload and the column resolved to false, which when a column name is actually configured, can only mean it isn't in the table. So a non-empty configured column now produces a clear "Input Image Column '' not found in the input table. Available columns: [...]" error instead of the generic "No image source" message. Same for audio.

Any related issues, documentation, discussions?

Closes #7197.

How was this PR tested?

sbt "WorkflowOperator/testOnly org.apache.texera.amber.operator.huggingFace.* org.apache.texera.amber.util.PythonCodeRawInvalidTextSpec": passes (126 tests). Added a test asserting the generated script produces the "Input Image/Audio Column '...' not found" errors; PythonCodeRawInvalidTextSpec py-compiles the generated Python (guards the added nested indentation). scalafmt clean.

Was this PR authored or co-authored using generative AI tooling?

This PR was co-authored with Claude Opus 4.8 in compliance with ASF policy.

[github-actions]

Backport auto-label report

This fix: PR was checked against each actively-supported release branch. release/* labels drive the post-merge backport, so add or remove one to change where this fix lands.

Release branch

Analysis

⚠️ release/v1.2

Not labeled automatically — none of the files this PR modifies exist on this branch (common/workflow-operator/src/main/scala/org/apache/texera/amber/operator/huggingFace/codegen/HuggingFaceCodegenBase.scala, common/workflow-operator/src/test/scala/org/apache/texera/amber/operator/huggingFace/HuggingFaceInferenceOpDescSpec.scala). The fix may target code that isn't on this release, or the files were moved/renamed after the branch was cut. Please check and add release/v1.2 by hand if this fix should be backported here.

Auto-label run.

[github-actions]

Automated Reviewer Suggestions

Based on the git blame history of the changed files, we recommend the following reviewers:

• No candidates found from git blame history.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 85.44%. Comparing base (114a610) to head (88dacb0).

Additional details and impacted files

@@            Coverage Diff            @@
##               main    #7299   +/-   ##
=========================================
  Coverage     85.43%   85.44%           
  Complexity     4150     4150           
=========================================
  Files          1169     1169           
  Lines         46740    46740           
  Branches       5202     5202           
=========================================
+ Hits          39934    39936    +2     
+ Misses         5092     5090    -2     
  Partials       1714     1714           

Flag	Coverage Δ		*Carryforward flag
access-control-service	70.00% <ø> (ø)
agent-service	85.50% <ø> (ø)		Carriedforward from e30eb17
amber	80.88% <ø> (ø)
computing-unit-managing-service	50.72% <ø> (ø)
config-service	65.97% <ø> (ø)
file-service	69.05% <ø> (ø)
frontend	88.11% <ø> (ø)		Carriedforward from e30eb17
notebook-migration-service	78.89% <ø> (ø)
pyamber	97.55% <ø> (+0.04%)	⬆️	Carriedforward from e30eb17
workflow-compiling-service	26.31% <ø> (ø)

*This pull request uses carry forward flags. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[github-actions]

⚠️ Benchmark changes need a look

🟢 2 better · 🔴 6 worse · ⚪ 7 noise (<±5%) · 0 without baseline

Compared against main 114a610 benchmarked on this same runner, so the delta is largely free of cross-runner hardware noise. The "7d avg" column still reflects the gh-pages dashboard. Treat <±5% as noise unless repeated.

Dashboard · Run

	config	throughput	MB/s	latency	max Δ latest / 7d
🔴	bs=10 sw=10 sl=64	420	0.256	23,223/30,296/30,296 us	🔴 +12.7% / 🔴 +84.8%
🔴	bs=100 sw=10 sl=64	889	0.543	111,060/140,933/140,933 us	🟢 -12.2% / 🔴 +25.3%
🔴	bs=1000 sw=10 sl=64	1,077	0.657	912,376/1,074,213/1,074,213 us	🔴 +11.9% / 🟢 -11.8%

Baseline details

Latest main 114a610 from same runner

config	metric	PR	latest main	7d avg	Δ latest	Δ 7d
bs=10 sw=10 sl=64	throughput	420 tuples/sec	471 tuples/sec	743.73 tuples/sec	-10.8%	-43.5%
bs=10 sw=10 sl=64	MB/s	0.256 MB/s	0.287 MB/s	0.454 MB/s	-10.8%	-43.6%
bs=10 sw=10 sl=64	p50	23,223 us	20,613 us	13,130 us	+12.7%	+76.9%
bs=10 sw=10 sl=64	p95	30,296 us	31,558 us	16,391 us	-4.0%	+84.8%
bs=10 sw=10 sl=64	p99	30,296 us	31,558 us	19,408 us	-4.0%	+56.1%
bs=100 sw=10 sl=64	throughput	889 tuples/sec	886 tuples/sec	944.04 tuples/sec	+0.3%	-5.8%
bs=100 sw=10 sl=64	MB/s	0.543 MB/s	0.541 MB/s	0.576 MB/s	+0.4%	-5.8%
bs=100 sw=10 sl=64	p50	111,060 us	104,818 us	105,853 us	+6.0%	+4.9%
bs=100 sw=10 sl=64	p95	140,933 us	160,585 us	112,493 us	-12.2%	+25.3%
bs=100 sw=10 sl=64	p99	140,933 us	160,585 us	122,200 us	-12.2%	+15.3%
bs=1000 sw=10 sl=64	throughput	1,077 tuples/sec	1,108 tuples/sec	972.22 tuples/sec	-2.8%	+10.8%
bs=1000 sw=10 sl=64	MB/s	0.657 MB/s	0.676 MB/s	0.593 MB/s	-2.8%	+10.7%
bs=1000 sw=10 sl=64	p50	912,376 us	896,043 us	1,033,856 us	+1.8%	-11.8%
bs=1000 sw=10 sl=64	p95	1,074,213 us	959,779 us	1,076,083 us	+11.9%	-0.2%
bs=1000 sw=10 sl=64	p99	1,074,213 us	959,779 us	1,107,701 us	+11.9%	-3.0%

Raw CSV

config_idx,batch_size,schema_width,string_len,num_batches,total_ms,total_tuples,total_bytes,tuples_per_sec,mb_per_sec,lat_p50_us,lat_p95_us,lat_p99_us
0,10,10,64,20,475.98,200,128000,420,0.256,23223.36,30296.34,30296.34
1,100,10,64,20,2249.52,2000,1280000,889,0.543,111059.61,140932.73,140932.73
2,1000,10,64,20,18570.59,20000,12800000,1077,0.657,912375.68,1074212.68,1074212.68

[PG1204]

/request-review @xuang7

[chenlica]

@PG1204 Screenshots?

[PG1204]

@PG1204 Screenshots?

hi @chenlica
so there isn't a natural UI screenshot here, Input Image Column is an autofill dropdown, so the client only offers existing columns and clears the reference once its column disappears and a "configured-but-missing" column can't be held in the GUI. The improved error is a runtime message that surfaces when the operator executes with such a config (headless/API/scheduled runs, or a stored workflow whose upstream schema changed).

this is reproduced by running the operator's generated Python against an input missing the configured column (old message was the misleading "No image source…"):

[chenlica]

@PG1204 Any reasonable screenshot that can improve the readability of the PR will be appreciated. If you don't think that's doable, that's fine.

[PG1204]

@PG1204 Any reasonable screenshot that can improve the readability of the PR will be appreciated. If you don't think that's doable, that's fine.

@chenlica both sides are the operator's real generated Python run against the same input with the same config

Before/after of the error message for an image task (image-to-text) whose Input Image Column is set to a column that isn't in the input (img, while the input table has only prompt), with no image upload.

Before: Image task configuration error: No image source. Set an Input Image Column or upload an image. This is misleading, since a column was set.

After (this PR): Image task configuration error: Input Image Column 'img' not found in the input table. Available columns: ['prompt'], which names the missing column and lists the available ones.

Captured by running the operator's generated Python headlessly against a one-column table (this state is reached via API/headless/stored-workflow execution; the GUI clears out-of-schema column references, so it isn't reproducible by clicking through the editor).