fix(text): correct positions on duplicate markdown chunks #1120
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Because - The markdown splitter in the text component returned wrong chunk positions when one of the sub-chunks (chunks before merge) had a content that occrred several times in the raw text. Additionally, after merging, that sub-chunk had to end up the last in the merged chunk, which caused the end position to be wrong. This commit - Uses the block positions as the starting point for the chunk location. - Adds a test with the document that signaled the error.
jvallesm
force-pushed
the
jvalles/fix-chunk-positions
branch
from
e2f53b3 to
8ec6ccd
Compare
This commit - Applies several code health improvements to the markdown chunking code, such as: - Clearer and conciser function and code comments - More idiomatic loops using the range function - Reduced nesting in if/else statements
This was referenced Sep 22, 2025
jvallesm
added a commit
to instill-ai/instill-core
that referenced
this pull request
Sep 23, 2025
Because - The version of the pipeline-backend service is not updated in the instill-core repository. This commit - updates the `PIPELINE_BACKEND_VERSION` in the `.env` file to `1b4cd1f`. - updates the `pipelineBackend.image.tag` in the helm chart values.yaml file to `1b4cd1f`. ## Changes in pipeline-backend - fix(text): correct positions on duplicate markdown chunks (instill-ai/pipeline-backend#1120) - refactor(component,generic,http): replace env-based URL validation with constructor injection (instill-ai/pipeline-backend#1121) - fix(usage): add missing error filtering for users/admin (instill-ai/pipeline-backend#1119) - feat(component,ai,gemini): implement File API support for large files… (instill-ai/pipeline-backend#1118) - perf(data): optimize struct marshaling/unmarshaling with caching and … (instill-ai/pipeline-backend#1117) - feat(data): enhance unmarshaler with JSON string to struct conversion (instill-ai/pipeline-backend#1116) - feat(data): implement time types support with pattern validation (instill-ai/pipeline-backend#1115) - feat(component,ai,gemini): add multimedia support with unified format… (instill-ai/pipeline-backend#1114) - ci(workflows): adopt GitHub-hosted runner (instill-ai/pipeline-backend#1113) - perf(data): enhance comprehensive format coverage and optimize test performance (instill-ai/pipeline-backend#1112) - ci(workflows): adopt loarger runner for coverage test (instill-ai/pipeline-backend#1111) - perf(component,operator,document): optimize unit tests and fix LibreOffice dependency failures (instill-ai/pipeline-backend#1110) - perf(component,operator,video): optimize unit test performance by 59.7% (instill-ai/pipeline-backend#1109) - perf(component,operator,image): optimize unit tests for 98.5% faster … (instill-ai/pipeline-backend#1107) - ci(docker): optimize Dockerfiles with multi-stage builds for faster build times (instill-ai/pipeline-backend#1108) - perf(data): implement automatic field naming convention detection with LRU caching (instill-ai/pipeline-backend#1105) - feat(component,ai,gemini): enhance streaming to output all fields (instill-ai/pipeline-backend#1106) - fix(component,ai,gemini): correct text-based documents logic (instill-ai/pipeline-backend#1103) - test(component,generic,http): replace external httpbin.org dependency with local test server (instill-ai/pipeline-backend#1101) - ci(docker): add GitHub fallback for ffmpeg installation (instill-ai/pipeline-backend#1102) Co-authored-by: jvallesm <3977183+jvallesm@users.noreply.github.com>
jvallesm
pushed a commit
that referenced
this pull request
Oct 7, 2025
🤖 I have created a release *beep* *boop* --- ## [0.61.0](v0.60.0...v0.61.0) (2025-10-06) ### Features * **component,ai,gemini:** add image generation support ([#1122](#1122)) ([d986614](d986614)) * **component,ai,gemini:** add multimedia support with unified format… ([#1114](#1114)) ([291b379](291b379)) * **component,ai,gemini:** add text embeddings task support ([#1129](#1129)) ([d7ca6cf](d7ca6cf)) * **component,ai,gemini:** enhance streaming to output all fields ([#1106](#1106)) ([dfb6b24](dfb6b24)) * **component,ai,gemini:** implement automatic format conversion for unsupported media types ([#1128](#1128)) ([f767b8a](f767b8a)) * **component,ai,gemini:** implement File API support for large files… ([#1118](#1118)) ([b51c8f4](b51c8f4)) * **data:** add comprehensive AVIF image format support ([#1135](#1135)) ([76d6941](76d6941)) * **data:** add HEIC/HEIF image support and normalize MIME types ([#1127](#1127)) ([2dfa254](2dfa254)) * **data:** enhance unmarshaler with JSON string to struct conversion ([#1116](#1116)) ([9e06b7c](9e06b7c)) * **data:** implement time types support with pattern validation ([#1115](#1115)) ([79630c0](79630c0)) ### Bug Fixes * **compogen:** escape curly braces for readme.com compatibility ([#1124](#1124)) ([904992d](904992d)) * **component,ai,gemini:** add operation validation for cache task ([#1130](#1130)) ([9e19255](9e19255)) * **component,ai,gemini:** correct text-based documents logic ([#1103](#1103)) ([ed5a111](ed5a111)) * **component,ai,gemini:** unify InlineData processing and enable images in streaming responses ([#1125](#1125)) ([3117046](3117046)) * **data:** remove duplicate dot in generated filenames ([#1136](#1136)) ([0a74a00](0a74a00)) * **external:** fix Content-Disposition header parsing for filename extraction ([#1132](#1132)) ([869b081](869b081)) * **service:** handle null JSON metadata in pipeline conversion ([#1134](#1134)) ([b244784](b244784)) * **text:** correct positions on duplicate markdown chunks ([#1120](#1120)) ([1b4cd1f](1b4cd1f)) * **usage:** add missing error filtering for users/admin ([#1119](#1119)) ([cd1bd55](cd1bd55)) ### Refactor * **component,ai,gemini:** merge usage and usage-metadata fields into single usage field ([#1126](#1126)) ([a6046cd](a6046cd)) * **component,ai.gemini:** standardize file api timeout and use native embedding type ([#1133](#1133)) ([174f7d6](174f7d6)) * **component,generic,http:** move test functions to test files and improve code legibility ([#1131](#1131)) ([1153a09](1153a09)) * **component,generic,http:** replace env-based URL validation with constructor injection ([#1121](#1121)) ([f1f7d2f](f1f7d2f)) ### Tests * **component,generic,http:** replace external httpbin.org dependency with local test server ([#1101](#1101)) ([a82d155](a82d155)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).
donch1989
pushed a commit
that referenced
this pull request
Oct 7, 2025
🤖 I have created a release *beep* *boop* --- ## [0.61.0](v0.60.0...v0.61.0) (2025-10-07) ### Features * **component,ai,gemini:** add image generation support ([#1122](#1122)) ([d986614](d986614)) * **component,ai,gemini:** add multimedia support with unified format… ([#1114](#1114)) ([291b379](291b379)) * **component,ai,gemini:** add text embeddings task support ([#1129](#1129)) ([d7ca6cf](d7ca6cf)) * **component,ai,gemini:** enhance streaming to output all fields ([#1106](#1106)) ([dfb6b24](dfb6b24)) * **component,ai,gemini:** implement automatic format conversion for unsupported media types ([#1128](#1128)) ([f767b8a](f767b8a)) * **component,ai,gemini:** implement File API support for large files… ([#1118](#1118)) ([b51c8f4](b51c8f4)) * **data:** add comprehensive AVIF image format support ([#1135](#1135)) ([76d6941](76d6941)) * **data:** add HEIC/HEIF image support and normalize MIME types ([#1127](#1127)) ([2dfa254](2dfa254)) * **data:** enhance unmarshaler with JSON string to struct conversion ([#1116](#1116)) ([9e06b7c](9e06b7c)) * **data:** implement time types support with pattern validation ([#1115](#1115)) ([79630c0](79630c0)) ### Bug Fixes * **compogen:** escape curly braces for readme.com compatibility ([#1124](#1124)) ([904992d](904992d)) * **component,ai,gemini:** add operation validation for cache task ([#1130](#1130)) ([9e19255](9e19255)) * **component,ai,gemini:** correct text-based documents logic ([#1103](#1103)) ([ed5a111](ed5a111)) * **component,ai,gemini:** unify InlineData processing and enable images in streaming responses ([#1125](#1125)) ([3117046](3117046)) * **component,document:** fix incorrect expected value in the unit test ([#1138](#1138)) ([189dbd6](189dbd6)) * **data:** remove duplicate dot in generated filenames ([#1136](#1136)) ([0a74a00](0a74a00)) * **external:** fix Content-Disposition header parsing for filename extraction ([#1132](#1132)) ([869b081](869b081)) * **service:** handle null JSON metadata in pipeline conversion ([#1134](#1134)) ([b244784](b244784)) * **text:** correct positions on duplicate markdown chunks ([#1120](#1120)) ([1b4cd1f](1b4cd1f)) * **usage:** add missing error filtering for users/admin ([#1119](#1119)) ([cd1bd55](cd1bd55)) ### Miscellaneous * release v0.61.0 ([e1db93c](e1db93c)) ### Refactor * **component,ai,gemini:** merge usage and usage-metadata fields into single usage field ([#1126](#1126)) ([a6046cd](a6046cd)) * **component,ai.gemini:** standardize file api timeout and use native embedding type ([#1133](#1133)) ([174f7d6](174f7d6)) * **component,generic,http:** move test functions to test files and improve code legibility ([#1131](#1131)) ([1153a09](1153a09)) * **component,generic,http:** replace env-based URL validation with constructor injection ([#1121](#1121)) ([f1f7d2f](f1f7d2f)) ### Tests * **component,generic,http:** replace external httpbin.org dependency with local test server ([#1101](#1101)) ([a82d155](a82d155)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Because
positions when one of the sub-chunks (chunks before merge) had a
content that occrred several times in the raw text. Additionally,
after merging, that sub-chunk had to end up the last in the merged
chunk, which caused the end position to be wrong.
This commit