Skip to content

Commit caad330

Browse files
Google APIscopybara-github
authored andcommitted
feat: Support Chunk header and footer in Doc AI external proto
docs: Keep the API doc up-to-date with recent changes PiperOrigin-RevId: 627100297
1 parent 5fa62a9 commit caad330

1 file changed

Lines changed: 29 additions & 4 deletions

File tree

google/cloud/documentai/v1beta3/document.proto

Lines changed: 29 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -286,16 +286,16 @@ message Document {
286286
// Whether the text is underlined.
287287
bool underlined = 7;
288288

289-
// Whether the text is strikethrough.
289+
// Whether the text is strikethrough. This feature is not supported yet.
290290
bool strikeout = 8;
291291

292-
// Whether the text is a subscript.
292+
// Whether the text is a subscript. This feature is not supported yet.
293293
bool subscript = 9;
294294

295-
// Whether the text is a superscript.
295+
// Whether the text is a superscript. This feature is not supported yet.
296296
bool superscript = 10;
297297

298-
// Whether the text is in small caps.
298+
// Whether the text is in small caps. This feature is not supported yet.
299299
bool smallcaps = 11;
300300

301301
// TrueType weight on a scale `100` (thin) to `1000` (ultra-heavy).
@@ -1010,9 +1010,28 @@ message Document {
10101010
int32 page_end = 2;
10111011
}
10121012

1013+
// Represents the page header associated with the chunk.
1014+
message ChunkPageHeader {
1015+
// Header in text format.
1016+
string text = 1;
1017+
1018+
// Page span of the header.
1019+
ChunkPageSpan page_span = 2;
1020+
}
1021+
1022+
// Represents the page footer associated with the chunk.
1023+
message ChunkPageFooter {
1024+
// Footer in text format.
1025+
string text = 1;
1026+
1027+
// Page span of the footer.
1028+
ChunkPageSpan page_span = 2;
1029+
}
1030+
10131031
// ID of the chunk.
10141032
string chunk_id = 1;
10151033

1034+
// DO NOT USE.
10161035
// List of all parsed documents layout source blocks used to generate the
10171036
// chunk.
10181037
repeated string source_block_ids = 2;
@@ -1022,6 +1041,12 @@ message Document {
10221041

10231042
// Page span of the chunk.
10241043
ChunkPageSpan page_span = 4;
1044+
1045+
// Page headers associated with the chunk.
1046+
repeated ChunkPageHeader page_headers = 5;
1047+
1048+
// Page footers associated with the chunk.
1049+
repeated ChunkPageFooter page_footers = 6;
10251050
}
10261051

10271052
// List of chunks.

0 commit comments

Comments
 (0)