docling 2.1.0__py3-none-any.whl → 2.2.1__py3-none-any.whl

docling/backend/abstract_backend.py

@@ -13,6 +13,7 @@ if TYPE_CHECKING:
 class AbstractDocumentBackend(ABC):
     @abstractmethod
     def __init__(self, in_doc: "InputDocument", path_or_stream: Union[BytesIO, Path]):
+        self.file = in_doc.file
         self.path_or_stream = path_or_stream
         self.document_hash = in_doc.document_hash
         self.input_format = in_doc.format

docling/backend/asciidoc_backend.py

@@ -0,0 +1,435 @@
+import logging
+import os
+import re
+from io import BytesIO
+from pathlib import Path
+from typing import Set, Union
+
+from docling_core.types.doc import (
+    DocItem,
+    DocItemLabel,
+    DoclingDocument,
+    DocumentOrigin,
+    GroupItem,
+    GroupLabel,
+    ImageRef,
+    NodeItem,
+    Size,
+    TableCell,
+    TableData,
+)
+from pydantic import AnyUrl
+
+from docling.backend.abstract_backend import DeclarativeDocumentBackend
+from docling.datamodel.base_models import InputFormat
+from docling.datamodel.document import InputDocument
+
+_log = logging.getLogger(__name__)
+
+
+class AsciiDocBackend(DeclarativeDocumentBackend):
+
+    def __init__(self, in_doc: InputDocument, path_or_stream: Union[BytesIO, Path]):
+        super().__init__(in_doc, path_or_stream)
+
+        self.path_or_stream = path_or_stream
+
+        try:
+            if isinstance(self.path_or_stream, BytesIO):
+                text_stream = self.path_or_stream.getvalue().decode("utf-8")
+                self.lines = text_stream.split("\n")
+            if isinstance(self.path_or_stream, Path):
+                with open(self.path_or_stream, "r", encoding="utf-8") as f:
+                    self.lines = f.readlines()
+            self.valid = True
+
+        except Exception as e:
+            raise RuntimeError(
+                f"Could not initialize AsciiDoc backend for file with hash {self.document_hash}."
+            ) from e
+        return
+
+    def is_valid(self) -> bool:
+        return self.valid
+
+    @classmethod
+    def supports_pagination(cls) -> bool:
+        return False
+
+    def unload(self):
+        return
+
+    @classmethod
+    def supported_formats(cls) -> Set[InputFormat]:
+        return {InputFormat.ASCIIDOC}
+
+    def convert(self) -> DoclingDocument:
+        """
+        Parses the ASCII into a structured document model.
+        """
+
+        origin = DocumentOrigin(
+            filename=self.file.name or "file",
+            mimetype="text/asciidoc",
+            binary_hash=self.document_hash,
+        )
+
+        doc = DoclingDocument(name=self.file.stem or "file", origin=origin)
+
+        doc = self._parse(doc)
+
+        return doc
+
+    def _parse(self, doc: DoclingDocument):
+        """
+        Main function that orchestrates the parsing by yielding components:
+        title, section headers, text, lists, and tables.
+        """
+
+        content = ""
+
+        in_list = False
+        in_table = False
+
+        text_data: list[str] = []
+        table_data: list[str] = []
+        caption_data: list[str] = []
+
+        # parents: dict[int, Union[DocItem, GroupItem, None]] = {}
+        parents: dict[int, Union[GroupItem, None]] = {}
+        # indents: dict[int, Union[DocItem, GroupItem, None]] = {}
+        indents: dict[int, Union[GroupItem, None]] = {}
+
+        for i in range(0, 10):
+            parents[i] = None
+            indents[i] = None
+
+        for line in self.lines:
+            # line = line.strip()
+
+            # Title
+            if self._is_title(line):
+                item = self._parse_title(line)
+                level = item["level"]
+
+                parents[level] = doc.add_text(
+                    text=item["text"], label=DocItemLabel.TITLE
+                )
+
+            # Section headers
+            elif self._is_section_header(line):
+                item = self._parse_section_header(line)
+                level = item["level"]
+
+                parents[level] = doc.add_heading(
+                    text=item["text"], level=item["level"], parent=parents[level - 1]
+                )
+                for k, v in parents.items():
+                    if k > level:
+                        parents[k] = None
+
+            # Lists
+            elif self._is_list_item(line):
+
+                _log.debug(f"line: {line}")
+                item = self._parse_list_item(line)
+                _log.debug(f"parsed list-item: {item}")
+
+                level = self._get_current_level(parents)
+
+                if not in_list:
+                    in_list = True
+
+                    parents[level + 1] = doc.add_group(
+                        parent=parents[level], name="list", label=GroupLabel.LIST
+                    )
+                    indents[level + 1] = item["indent"]
+
+                elif in_list and item["indent"] > indents[level]:
+                    parents[level + 1] = doc.add_group(
+                        parent=parents[level], name="list", label=GroupLabel.LIST
+                    )
+                    indents[level + 1] = item["indent"]
+
+                elif in_list and item["indent"] < indents[level]:
+
+                    # print(item["indent"], " => ", indents[level])
+                    while item["indent"] < indents[level]:
+                        # print(item["indent"], " => ", indents[level])
+                        parents[level] = None
+                        indents[level] = None
+                        level -= 1
+
+                doc.add_list_item(
+                    item["text"], parent=self._get_current_parent(parents)
+                )
+
+            elif in_list and not self._is_list_item(line):
+                in_list = False
+
+                level = self._get_current_level(parents)
+                parents[level] = None
+
+            # Tables
+            elif line.strip() == "|===" and not in_table:  # start of table
+                in_table = True
+
+            elif self._is_table_line(line):  # within a table
+                in_table = True
+                table_data.append(self._parse_table_line(line))
+
+            elif in_table and (
+                (not self._is_table_line(line)) or line.strip() == "|==="
+            ):  # end of table
+
+                caption = None
+                if len(caption_data) > 0:
+                    caption = doc.add_text(
+                        text=" ".join(caption_data), label=DocItemLabel.CAPTION
+                    )
+
+                caption_data = []
+
+                data = self._populate_table_as_grid(table_data)
+                doc.add_table(
+                    data=data, parent=self._get_current_parent(parents), caption=caption
+                )
+
+                in_table = False
+                table_data = []
+
+            # Picture
+            elif self._is_picture(line):
+
+                caption = None
+                if len(caption_data) > 0:
+                    caption = doc.add_text(
+                        text=" ".join(caption_data), label=DocItemLabel.CAPTION
+                    )
+
+                caption_data = []
+
+                item = self._parse_picture(line)
+
+                size = None
+                if "width" in item and "height" in item:
+                    size = Size(width=int(item["width"]), height=int(item["height"]))
+
+                uri = None
+                if (
+                    "uri" in item
+                    and not item["uri"].startswith("http")
+                    and item["uri"].startswith("//")
+                ):
+                    uri = "file:" + item["uri"]
+                elif (
+                    "uri" in item
+                    and not item["uri"].startswith("http")
+                    and item["uri"].startswith("/")
+                ):
+                    uri = "file:/" + item["uri"]
+                elif "uri" in item and not item["uri"].startswith("http"):
+                    uri = "file://" + item["uri"]
+
+                image = ImageRef(mimetype="image/png", size=size, dpi=70, uri=uri)
+                doc.add_picture(image=image, caption=caption)
+
+            # Caption
+            elif self._is_caption(line) and len(caption_data) == 0:
+                item = self._parse_caption(line)
+                caption_data.append(item["text"])
+
+            elif (
+                len(line.strip()) > 0 and len(caption_data) > 0
+            ):  # allow multiline captions
+                item = self._parse_text(line)
+                caption_data.append(item["text"])
+
+            # Plain text
+            elif len(line.strip()) == 0 and len(text_data) > 0:
+                doc.add_text(
+                    text=" ".join(text_data),
+                    label=DocItemLabel.PARAGRAPH,
+                    parent=self._get_current_parent(parents),
+                )
+                text_data = []
+
+            elif len(line.strip()) > 0:  # allow multiline texts
+
+                item = self._parse_text(line)
+                text_data.append(item["text"])
+
+        if len(text_data) > 0:
+            doc.add_text(
+                text=" ".join(text_data),
+                label=DocItemLabel.PARAGRAPH,
+                parent=self._get_current_parent(parents),
+            )
+            text_data = []
+
+        if in_table and len(table_data) > 0:
+            data = self._populate_table_as_grid(table_data)
+            doc.add_table(data=data, parent=self._get_current_parent(parents))
+
+            in_table = False
+            table_data = []
+
+        return doc
+
+    def _get_current_level(self, parents):
+        for k, v in parents.items():
+            if v == None and k > 0:
+                return k - 1
+
+        return 0
+
+    def _get_current_parent(self, parents):
+        for k, v in parents.items():
+            if v == None and k > 0:
+                return parents[k - 1]
+
+        return None
+
+    #   =========   Title
+    def _is_title(self, line):
+        return re.match(r"^= ", line)
+
+    def _parse_title(self, line):
+        return {"type": "title", "text": line[2:].strip(), "level": 0}
+
+    #   =========   Section headers
+    def _is_section_header(self, line):
+        return re.match(r"^==+", line)
+
+    def _parse_section_header(self, line):
+        match = re.match(r"^(=+)\s+(.*)", line)
+
+        marker = match.group(1)  # The list marker (e.g., "*", "-", "1.")
+        text = match.group(2)  # The actual text of the list item
+
+        header_level = marker.count("=")  # number of '=' represents level
+        return {
+            "type": "header",
+            "level": header_level - 1,
+            "text": text.strip(),
+        }
+
+    #   =========   Lists
+    def _is_list_item(self, line):
+        return re.match(r"^(\s)*(\*|-|\d+\.|\w+\.) ", line)
+
+    def _parse_list_item(self, line):
+        """Extract the item marker (number or bullet symbol) and the text of the item."""
+
+        match = re.match(r"^(\s*)(\*|-|\d+\.)\s+(.*)", line)
+        if match:
+            indent = match.group(1)
+            marker = match.group(2)  # The list marker (e.g., "*", "-", "1.")
+            text = match.group(3)  # The actual text of the list item
+
+            if marker == "*" or marker == "-":
+                return {
+                    "type": "list_item",
+                    "marker": marker,
+                    "text": text.strip(),
+                    "numbered": False,
+                    "indent": 0 if indent == None else len(indent),
+                }
+            else:
+                return {
+                    "type": "list_item",
+                    "marker": marker,
+                    "text": text.strip(),
+                    "numbered": True,
+                    "indent": 0 if indent == None else len(indent),
+                }
+        else:
+            # Fallback if no match
+            return {
+                "type": "list_item",
+                "marker": "-",
+                "text": line,
+                "numbered": False,
+                "indent": 0,
+            }
+
+    #   =========   Tables
+    def _is_table_line(self, line):
+        return re.match(r"^\|.*\|", line)
+
+    def _parse_table_line(self, line):
+        # Split table cells and trim extra spaces
+        return [cell.strip() for cell in line.split("|") if cell.strip()]
+
+    def _populate_table_as_grid(self, table_data):
+
+        num_rows = len(table_data)
+
+        # Adjust the table data into a grid format
+        num_cols = max(len(row) for row in table_data)
+
+        data = TableData(num_rows=num_rows, num_cols=num_cols, table_cells=[])
+        for row_idx, row in enumerate(table_data):
+            # Pad rows with empty strings to match column count
+            # grid.append(row + [''] * (max_cols - len(row)))
+
+            for col_idx, text in enumerate(row):
+                row_span = 1
+                col_span = 1
+
+                cell = TableCell(
+                    text=text,
+                    row_span=row_span,
+                    col_span=col_span,
+                    start_row_offset_idx=row_idx,
+                    end_row_offset_idx=row_idx + row_span,
+                    start_col_offset_idx=col_idx,
+                    end_col_offset_idx=col_idx + col_span,
+                    col_header=False,
+                    row_header=False,
+                )
+                data.table_cells.append(cell)
+
+        return data
+
+    #   =========   Pictures
+    def _is_picture(self, line):
+        return re.match(r"^image::", line)
+
+    def _parse_picture(self, line):
+        """
+        Parse an image macro, extracting its path and attributes.
+        Syntax: image::path/to/image.png[Alt Text, width=200, height=150, align=center]
+        """
+        mtch = re.match(r"^image::(.+)\[(.*)\]$", line)
+        if mtch:
+            picture_path = mtch.group(1).strip()
+            attributes = mtch.group(2).split(",")
+            picture_info = {"type": "picture", "uri": picture_path}
+
+            # Extract optional attributes (alt text, width, height, alignment)
+            if attributes:
+                picture_info["alt"] = attributes[0].strip() if attributes[0] else ""
+                for attr in attributes[1:]:
+                    key, value = attr.split("=")
+                    picture_info[key.strip()] = value.strip()
+
+            return picture_info
+
+        return {"type": "picture", "uri": line}
+
+    #   =========   Captions
+    def _is_caption(self, line):
+        return re.match(r"^\.(.+)", line)
+
+    def _parse_caption(self, line):
+        mtch = re.match(r"^\.(.+)", line)
+        if mtch:
+            text = mtch.group(1)
+            return {"type": "caption", "text": text}
+
+        return {"type": "caption", "text": ""}
+
+    #   =========   Plain text
+    def _parse_text(self, line):
+        return {"type": "text", "text": line.strip()}

docling/backend/docling_parse_backend.py

@@ -6,7 +6,7 @@ from typing import Iterable, List, Optional, Union
 
 import pypdfium2 as pdfium
 from docling_core.types.doc import BoundingBox, CoordOrigin, Size
-from docling_parse.docling_parse import pdf_parser
+from docling_parse.docling_parse import pdf_parser_v1
 from PIL import Image, ImageDraw
 from pypdfium2 import PdfPage
 
@@ -19,7 +19,7 @@ _log = logging.getLogger(__name__)
 
 class DoclingParsePageBackend(PdfPageBackend):
     def __init__(
-        self, parser: pdf_parser, document_hash: str, page_no: int, page_obj: PdfPage
+        self, parser: pdf_parser_v1, document_hash: str, page_no: int, page_obj: PdfPage
     ):
         self._ppage = page_obj
         parsed_page = parser.parse_pdf_from_key_on_page(document_hash, page_no)
@@ -192,7 +192,7 @@ class DoclingParseDocumentBackend(PdfDocumentBackend):
         super().__init__(in_doc, path_or_stream)
 
         self._pdoc = pdfium.PdfDocument(self.path_or_stream)
-        self.parser = pdf_parser()
+        self.parser = pdf_parser_v1()
 
         success = False
         if isinstance(self.path_or_stream, BytesIO):

docling/backend/docling_parse_v2_backend.py

@@ -26,9 +26,9 @@ class DoclingParseV2PageBackend(PdfPageBackend):
         self._ppage = page_obj
         parsed_page = parser.parse_pdf_from_key_on_page(document_hash, page_no)
 
-        self.valid = "pages" in parsed_page
+        self.valid = "pages" in parsed_page and len(parsed_page["pages"]) == 1
         if self.valid:
-            self._dpage = parsed_page["pages"][page_no]
+            self._dpage = parsed_page["pages"][0]
         else:
             _log.info(
                 f"An error occured when loading page {page_no} of document {document_hash}."
@@ -223,7 +223,15 @@ class DoclingParseV2DocumentBackend(PdfDocumentBackend):
             )
 
     def page_count(self) -> int:
-        return len(self._pdoc)  # To be replaced with docling-parse API
+        # return len(self._pdoc)  # To be replaced with docling-parse API
+
+        len_1 = len(self._pdoc)
+        len_2 = self.parser.number_of_pages(self.document_hash)
+
+        if len_1 != len_2:
+            _log.error(f"Inconsistent number of pages: {len_1}!={len_2}")
+
+        return len_2
 
     def load_page(self, page_no: int) -> DoclingParseV2PageBackend:
         return DoclingParseV2PageBackend(

docling/backend/html_backend.py

@@ -7,6 +7,7 @@ from bs4 import BeautifulSoup
 from docling_core.types.doc import (
     DocItemLabel,
     DoclingDocument,
+    DocumentOrigin,
     GroupLabel,
     TableCell,
     TableData,
@@ -66,7 +67,13 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
 
     def convert(self) -> DoclingDocument:
         # access self.path_or_stream to load stuff
-        doc = DoclingDocument(name="dummy")
+        origin = DocumentOrigin(
+            filename=self.file.name or "file",
+            mimetype="text/html",
+            binary_hash=self.document_hash,
+        )
+
+        doc = DoclingDocument(name=self.file.stem or "file", origin=origin)
         _log.debug("Trying to convert HTML...")
 
         if self.is_valid():
@@ -129,7 +136,6 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
     def get_direct_text(self, item):
         """Get the direct text of the <li> element (ignoring nested lists)."""
         text = item.find(string=True, recursive=False)
-
         if isinstance(text, str):
             return text.strip()
 
@@ -142,21 +148,20 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
         if isinstance(item, str):
             return [item]
 
-        result.append(self.get_direct_text(item))
-
-        try:
-            # Iterate over the children (and their text and tails)
-            for child in item:
-                try:
-                    # Recursively get the child's text content
-                    result.extend(self.extract_text_recursively(child))
-                except:
-                    pass
-        except:
-            _log.warn("item has no children")
-            pass
-
-        return " ".join(result)
+        if item.name not in ["ul", "ol"]:
+            try:
+                # Iterate over the children (and their text and tails)
+                for child in item:
+                    try:
+                        # Recursively get the child's text content
+                        result.extend(self.extract_text_recursively(child))
+                    except:
+                        pass
+            except:
+                _log.warn("item has no children")
+                pass
+
+        return "".join(result) + " "
 
     def handle_header(self, element, idx, doc):
         """Handles header tags (h1, h2, etc.)."""
@@ -175,11 +180,6 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
                 parent=self.parents[0], label=DocItemLabel.TITLE, text=text
             )
 
-        elif hlevel == self.level:
-            self.parents[hlevel] = doc.add_text(
-                parent=self.parents[hlevel - 1], label=label, text=text
-            )
-
         elif hlevel > self.level:
 
             # add invisible group
@@ -189,10 +189,6 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
                     label=GroupLabel.SECTION,
                     parent=self.parents[i - 1],
                 )
-
-            self.parents[hlevel] = doc.add_text(
-                parent=self.parents[hlevel - 1], label=label, text=text
-            )
             self.level = hlevel
 
         elif hlevel < self.level:
@@ -201,12 +197,14 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
             for key, val in self.parents.items():
                 if key > hlevel:
                     self.parents[key] = None
-
-            self.parents[hlevel] = doc.add_text(
-                parent=self.parents[hlevel - 1], label=label, text=text
-            )
             self.level = hlevel
 
+        self.parents[hlevel] = doc.add_heading(
+            parent=self.parents[hlevel - 1],
+            text=text,
+            level=hlevel,
+        )
+
     def handle_paragraph(self, element, idx, doc):
         """Handles paragraph tags (p)."""
         if element.text is None:
@@ -248,7 +246,12 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
 
         if nested_lists:
             name = element.name
-            text = self.get_direct_text(element)
+            # Text in list item can be hidden within hierarchy, hence
+            # we need to extract it recursively
+            text = self.extract_text_recursively(element)
+            # Flatten text, remove break lines:
+            text = text.replace("\n", "").replace("\r", "")
+            text = " ".join(text.split()).strip()
 
             marker = ""
             enumerated = False
@@ -256,14 +259,15 @@ class HTMLDocumentBackend(DeclarativeDocumentBackend):
                 marker = str(index_in_list)
                 enumerated = True
 
-            # create a list-item
-            self.parents[self.level + 1] = doc.add_list_item(
-                text=text,
-                enumerated=enumerated,
-                marker=marker,
-                parent=self.parents[self.level],
-            )
-            self.level += 1
+            if len(text) > 0:
+                # create a list-item
+                self.parents[self.level + 1] = doc.add_list_item(
+                    text=text,
+                    enumerated=enumerated,
+                    marker=marker,
+                    parent=self.parents[self.level],
+                )
+                self.level += 1
 
             self.walk(element, doc)