Block

A single structural element extracted from one page — one heading, one paragraph, or one table — before any chunking happens. result.blocks is a list of these, ordered as they appear in the document. For the concept and how blocks relate to chunks, see Blocks.

result = docslicer.parse_document("report.pdf")
 
for block in result.blocks:
    print(block.type, block.page_number, block.text[:80])

Fields

Field	Type	Description
`id`	`str`	Unique identifier for this block within the result.
`type`	`str`	The kind of content this block holds — `paragraph`, `heading`, `table`, `exhibits`, `navigation`, and so on. See Block types for the full vocabulary.
`text`	`str`	The block's text. For `table` blocks this is the serialised table, rendered per the chosen table representation.
`char_count`	`int`	Character count of `text`.
`page_number`	`int`	1-based physical page number the block appears on.
`page_label`	`str \| None`	Printed page label, e.g. `"iv"`, `"A-6"`. Differs from `page_number` when documents use Roman numerals, lettered appendices, or custom numbering. `None` when no label was detected.
`section`	`str`	Document region this block belongs to (`body`, `front_matter`, an appendix, …). See Sections for the full list.
`chunk_id`	`str \| None`	ID of the chunk this block was merged into. `None` when the block was not assigned to a chunk — for example when parsed with `chunking=False`.
`bbox`	`BBox \| None`	Bounding box of the block on the page. Populated for PDF and for HTML rendered via Playwright; `None` for DOCX, PPTX, and HTML parsed without Playwright.
`link_url`	`list[str]`	Unique URLs found within this block's text. Empty list when none were detected.
`table_ids`	`list[str]`	IDs of tables whose content appears in this block. Use these to look up the full `Table` object from `result.tables`; for a `table` block this points to its own table.
`extra`	`dict`	Additional pipeline fields requested via `extra_fields` on the parse call. Empty dict by default.

Unlike Chunk, a block has no heading, path, or token_count — those are chunk-level. Instead it carries type (what kind of element it is) and chunk_id (the chunk it ended up in).

Examples

Filter by type

block.type names the kind of element. Pull just the tables, or just the headings:

tables   = [b for b in result.blocks if b.type == "table"]
headings = [b for b in result.blocks if b.type == "heading"]

See Block types for every value.

Group blocks back into their chunks

block.chunk_id links each block to the chunk it was merged into:

from collections import defaultdict
 
blocks_by_chunk = defaultdict(list)
for block in result.blocks:
    blocks_by_chunk[block.chunk_id].append(block)

Collect exhibit links

link_url carries any URLs found in the block — handy for SEC filings, where the exhibit index links straight to each exhibit document:

for block in result.blocks:
    if block.type == "exhibits":
        for url in block.link_url:
            print(block.text, "→", url)

See Browsing SEC exhibits for a worked example.

Use extra fields

result = docslicer.parse_document("report.pdf", extra_fields=["is_bold", "font_size"])
 
for block in result.blocks:
    print(block.extra["font_size"], block.text[:80])

Methods

`to_dict()`

Serialize the block to a plain dict. All fields are included.

d = block.to_dict()

`Block.from_dict(d)`

Reconstruct a Block from a dict produced by to_dict().

block = Block.from_dict(d)