When `layout=False`: Adds spaces where the difference between the `x1` of one character and the `x0` of the next is greater than `x_tolerance`. (If `x_tolerance_ratio` is not `None`, the extractor uses a dynamic `x_tolerance` equal to `x_tolerance_ratio * previous_character["size"]`.) Adds newline characters where the difference between the `doctop` of one character and the `doctop` of the next is greater than `y_tolerance`.
When `layout=True` (*experimental feature*): Attempts to mimic the structural layout of the text on the page(s), using `x_density` and `y_density` to determine the minimum number of characters/newlines per "point," the PDF unit of measurement. Passing `line_dir_render="ttb"/"btt"/"ltr"/"rtl"` and/or `char_dir_render="ttb"/"btt"/"ltr"/"rtl"` will output the the lines/characters in a different direction than the default. All remaining `**kwargs` are passed to `.extract_words(...)` (see below), the first step in calculating the layout.
split_at_punctuation='!"&\'()*+,.:;<=>?@[\]^\`\{\|\}~'. Unless you set `expand_ligatures=False`, ligatures such as `fi` will be expanded into their constituent letters (e.g., `fi`). Passing `return_chars=True` will add, to each word dictionary, a list of its constituent characters, as a list in the `"chars"` field.|
|`.extract_text_lines(layout=False, strip=True, return_chars=True, **kwargs)`|*Experimental feature* that returns a list of dictionaries representing the lines of text on the page. The `strip` parameter works analogously to Python's `str.strip()` method, and returns `text` attributes without their surrounding whitespace. (Only relevant when `layout = True`.) Setting `return_chars` to `False` will exclude the individual character objects from the returned text-line dicts. The remaining `**kwargs` are those you would pass to `.extract_text(layout=True, ...)`.|
|`.search(pattern, regex=True, case=True, main_group=0, return_groups=True, return_chars=True, layout=False, **kwargs)`|*Experimental feature* that allows you to search a page's text, returning a list of all instances that match the query. For each instance, the response dictionary object contains the matching text, any regex group matches, the bounding box coordinates, and the char objects themselves. `pattern` can be a compiled regular expression, an uncompiled regular expression, or a non-regex string. If `regex` is `False`, the pattern is treated as a non-regex string. If `case` is `False`, the search is performed in a case-insensitive manner. Setting `main_group` restricts the results to a specific regex group within the `pattern` (default of `0` means the entire match). Setting `return_groups` and/or `return_chars` to `False` will exclude the lists of the matched regex groups and/or characters from being added (as `"groups"` and `"chars"` to the return dicts). The `layout` parameter operates as it does for `.extract_text(...)`. The remaining `**kwargs` are those you would pass to `.extract_text(layout=True, ...)`. __Note__: Zero-width and all-whitespace matches are discarded, because they (generally) have no explicit position on the page. |
|`.dedupe_chars(tolerance=1, extra_attrs=("fontname", "size"))`| Returns a version of the page with duplicate chars — those sharing the same text, positioning (within `tolerance` x/y), and `extra_attrs` as other characters — removed. (See [Issue #71](https://github.com/jsvine/pdfplumber/issues/71) to understand the motivation.)|
## Extracting tables
`pdfplumber`'s approach to table detection borrows heavily from [Anssi Nurminen's master's thesis](https://trepo.tuni.fi/bitstream/handle/123456789/21520/Nurminen.pdf?sequence=3), and is inspired by [Tabula](https://github.com/tabulapdf/tabula-extractor/issues/16). It works like this:
1. For any given PDF page, find the lines that are (a) explicitly defined and/or (b) implied by the alignment of words on the page.
2. Merge overlapping, or nearly-overlapping, lines.
3. Find the intersections of all those lines.
4. Find the most granular set of rectangles (i.e., cells) that use these intersections as their vertices.
5. Group contiguous cells into tables.
### Table-extraction methods
`pdfplumber.Page` objects can call the following table methods:
| Method | Description |
|--------|-------------|
|`.find_tables(table_settings={})`|Returns a list of `Table` objects. The `Table` object provides access to the `.cells`, `.rows`, `.columns`, and `.bbox` properties, as well as the `.extract(x_tolerance=3, y_tolerance=3)` method.|
|`.find_table(table_settings={})`|Similar to `.find_tables(...)`, but returns the *largest* table on the page, as a `Table` object. If multiple tables have the same size — as measured by the number of cells — this method returns the table closest to the top of the page.|
|`.extract_tables(table_settings={})`|Returns the text extracted from *all* tables found on the page, represented as a list of lists of lists, with the structure `table -> row -> cell`.|
|`.extract_table(table_settings={})`|Returns the text extracted from the *largest* table on the page (see `.find_table(...)` above), represented as a list of lists, with the structure `row -> cell`.|
|`.debug_tablefinder(table_settings={})`|Returns an instance of the `TableFinder` class, with access to the `.edges`, `.intersections`, `.cells`, and `.tables` properties.|
For example:
```python
pdf = pdfplumber.open("path/to/my.pdf")
page = pdf.pages[0]
page.extract_table()
```
[Click here for a more detailed example.](examples/notebooks/extract-table-ca-warn-report.ipynb)
### Table-extraction settings
By default, `extract_tables` uses the page's vertical and horizontal lines (or rectangle edges) as cell-separators. But the method is highly customizable via the `table_settings` argument. The possible settings, and their defaults:
```python
{
"vertical_strategy": "lines",
"horizontal_strategy": "lines",
"explicit_vertical_lines": [],
"explicit_horizontal_lines": [],
"snap_tolerance": 3,
"snap_x_tolerance": 3,
"snap_y_tolerance": 3,
"join_tolerance": 3,
"join_x_tolerance": 3,
"join_y_tolerance": 3,
"edge_min_length": 3,
"edge_min_length_prefilter": 1,
"min_words_vertical": 3,
"min_words_horizontal": 1,
"intersection_tolerance": 3,
"intersection_x_tolerance": 3,
"intersection_y_tolerance": 3,
"text_tolerance": 3,
"text_x_tolerance": 3,
"text_y_tolerance": 3,
"text_*": …, # See below
}
```
| Setting | Description |
|---------|-------------|
|`"vertical_strategy"`| Either `"lines"`, `"lines_strict"`, `"text"`, or `"explicit"`. See explanation below.|
|`"horizontal_strategy"`| Either `"lines"`, `"lines_strict"`, `"text"`, or `"explicit"`. See explanation below.|
|`"explicit_vertical_lines"`| A list of vertical lines that explicitly demarcate cells in the table. Can be used in combination with any of the strategies above. Items in the list should be either numbers — indicating the `x` coordinate of a line the full height of the page — or `line`/`rect`/`curve` objects.|
|`"explicit_horizontal_lines"`| A list of horizontal lines that explicitly demarcate cells in the table. Can be used in combination with any of the strategies above. Items in the list should be either numbers — indicating the `y` coordinate of a line the full height of the page — or `line`/`rect`/`curve` objects.|
|`"snap_tolerance"`, `"snap_x_tolerance"`, `"snap_y_tolerance"`| Parallel lines within `snap_tolerance` points will be "snapped" to the same horizontal or vertical position.|
|`"join_tolerance"`, `"join_x_tolerance"`, `"join_y_tolerance"`| Line segments on the same infinite line, and whose ends are within `join_tolerance` of one another, will be "joined" into a single line segment.|
|`"edge_min_length"`| Edges shorter than `edge_min_length` will be discarded before attempting to reconstruct the table.|
|`"edge_min_length_prefilter"`| Edges shorter than `edge_min_length_prefilter` will be discarded during initial edge filtering from the page. Lowering this value (e.g., to `0.5`) can help capture small dashed lines that might otherwise be filtered out.|
|`"min_words_vertical"`| When using `"vertical_strategy": "text"`, at least `min_words_vertical` words must share the same alignment.|
|`"min_words_horizontal"`| When using `"horizontal_strategy": "text"`, at least `min_words_horizontal` words must share the same alignment.|
|`"intersection_tolerance"`, `"intersection_x_tolerance"`, `"intersection_y_tolerance"`| When combining edges into cells, orthogonal edges must be within `intersection_tolerance` points to be considered intersecting.|
|`"text_*"`| All settings prefixed with `text_` are then used when extracting text from each discovered table. All possible arguments to `Page.extract_text(...)` are also valid here.|
|`"text_x_tolerance"`, `"text_y_tolerance"`| These `text_`-prefixed settings *also* apply to the table-identification algorithm when the `text` strategy is used. I.e., when that algorithm searches for words, it will expect the individual letters in each word to be no more than `text_x_tolerance`/`text_y_tolerance` points apart.|
### Table-extraction strategies
Both `vertical_strategy` and `horizontal_strategy` accept the following options:
| Strategy | Description |
|----------|-------------|
| `"lines"` | Use the page's graphical lines — including the sides of rectangle objects — as the borders of potential table-cells. |
| `"lines_strict"` | Use the page's graphical lines — but *not* the sides of rectangle objects — as the borders of potential table-cells. |
| `"text"` | For `vertical_strategy`: Deduce the (imaginary) lines that connect the left, right, or center of words on the page, and use those lines as the borders of potential table-cells. For `horizontal_strategy`, the same but using the tops of words. |
| `"explicit"` | Only use the lines explicitly defined in `explicit_vertical_lines` / `explicit_horizontal_lines`. |
### Notes
- Often it's helpful to crop a page — `Page.crop(bounding_box)` — before trying to extract the table.
- Table extraction for `pdfplumber` was radically redesigned for `v0.5.0`, and introduced breaking changes.
## Extracting form values
Sometimes PDF files can contain forms that include inputs that people can fill out and save. While values in form fields appear like other text in a PDF file, form data is handled differently. If you want the gory details, see page 671 of this [specification](https://opensource.adobe.com/dc-acrobat-sdk-docs/pdfstandards/pdfreference1.7old.pdf).
`pdfplumber` doesn't have an interface for working with form data, but you can access it using `pdfplumber`'s wrappers around `pdfminer`.
For example, this snippet will retrieve form field names and values and store them in a dictionary.
```python
import pdfplumber
from pdfplumber.utils.pdfinternals import resolve_and_decode, resolve
pdf = pdfplumber.open("document_with_form.pdf")
def parse_field_helper(form_data, field, prefix=None):
""" appends any PDF AcroForm field/value pairs in `field` to provided `form_data` list
if `field` has child fields, those will be parsed recursively.
"""
resolved_field = field.resolve()
field_name = '.'.join(filter(lambda x: x, [prefix, resolve_and_decode(resolved_field.get("T"))]))
if "Kids" in resolved_field:
for kid_field in resolved_field["Kids"]:
parse_field_helper(form_data, kid_field, prefix=field_name)
if "T" in resolved_field or "TU" in resolved_field:
# "T" is a field-name, but it's sometimes absent.
# "TU" is the "alternate field name" and is often more human-readable
# your PDF may have one, the other, or both.
alternate_field_name = resolve_and_decode(resolved_field.get("TU")) if resolved_field.get("TU") else None
field_value = resolve_and_decode(resolved_field["V"]) if 'V' in resolved_field else None
form_data.append([field_name, alternate_field_name, field_value])
form_data = []
fields = resolve(resolve(pdf.doc.catalog["AcroForm"])["Fields"])
for field in fields:
parse_field_helper(form_data, field)
```
Once you run this script, `form_data` is a list containing a three-element tuple for each form element. For instance, a PDF form with a city and state field might look like this.
```
[
['STATE.0', 'enter STATE', 'CA'],
['section 2 accident infoRmation.1.0',
'enter city of accident',
'SAN FRANCISCO']
]
```
*Thanks to [@jeremybmerrill](https://github.com/jeremybmerrill) for helping to maintain the form-parsing code above.*
## Demonstrations
- [Using `extract_table` on a California Worker Adjustment and Retraining Notification (WARN) report](examples/notebooks/extract-table-ca-warn-report.ipynb). Demonstrates basic visual debugging and table extraction.
- [Using `extract_table` on the FBI's National Instant Criminal Background Check System PDFs](examples/notebooks/extract-table-nics.ipynb). Demonstrates how to use visual debugging to find optimal table extraction settings. Also demonstrates `Page.crop(...)` and `Page.extract_text(...).`
- [Inspecting and visualizing `curve` objects](examples/notebooks/ag-energy-roundup-curves.ipynb).
- [Extracting fixed-width data from a San Jose PD firearm search report](examples/notebooks/san-jose-pd-firearm-report.ipynb), an example of using `Page.extract_text(...)`.
## Comparison to other libraries
Several other Python libraries help users to extract information from PDFs. As a broad overview, `pdfplumber` distinguishes itself from other PDF processing libraries by combining these features:
- Easy access to detailed information about each PDF object
- Higher-level, customizable methods for extracting text and tables
- Tightly integrated visual debugging
- Other useful utility functions, such as filtering objects via a crop-box
It's also helpful to know what features `pdfplumber` does __not__ provide:
- PDF *generation*
- PDF *modification*
- Optical character recognition (OCR)
- Strong support for extracting tables from OCR'ed documents
### Specific comparisons
- [`pdfminer.six`](https://github.com/pdfminer/pdfminer.six) provides the foundation for `pdfplumber`. It primarily focuses on parsing PDFs, analyzing PDF layouts and object positioning, and extracting text. It does not provide tools for table extraction or visual debugging. License: [MIT](https://github.com/pdfminer/pdfminer.six?tab=MIT-1-ov-file).
- [`PyPDF2`](https://github.com/mstamy2/PyPDF2) is a pure-Python library "capable of splitting, merging, cropping, and transforming the pages of PDF files. It can also add custom data, viewing options, and passwords to PDF files." It can extract page text, but does not provide easy access to shape objects (rectangles, lines, etc.), table-extraction, or visually debugging tools. License: [BSD](https://github.com/py-pdf/pypdf?tab=License-1-ov-file#readme).
- [`pymupdf`](https://pymupdf.readthedocs.io/) is substantially faster than `pdfminer.six` (and thus also `pdfplumber`) and can generate and modify PDFs, but the library requires installation of non-Python software (MuPDF). It also does not enable easy access to shape objects (rectangles, lines, etc.), and does not provide table-extraction or visual debugging tools. License: [AGPL](https://pymupdf.readthedocs.io/en/latest/about.html#license-and-copyright).
- [`camelot`](https://github.com/camelot-dev/camelot), [`tabula-py`](https://github.com/chezou/tabula-py), and [`pdftables`](https://github.com/drj11/pdftables) all focus primarily on extracting tables. In some cases, they may be better suited to the particular tables you are trying to extract. License: [MIT](https://github.com/camelot-dev/camelot?tab=MIT-1-ov-file#readme) (`camelot`), [MIT](https://github.com/chezou/tabula-py?tab=MIT-1-ov-file#readme) (`tabula-py`), [BSD](https://github.com/drj11/pdftables?tab=BSD-2-Clause-1-ov-file#readme) (`pdftables`).
## Acknowledgments / Contributors
Many thanks to the following users who've contributed ideas, features, and fixes:
- [Jacob Fenton](https://github.com/jsfenfen)
- [Dan Nguyen](https://github.com/dannguyen)
- [Jeff Barrera](https://github.com/jeffbarrera)
- [Bob Lannon](https://github.com/boblannon)
- [Dustin Tindall](https://github.com/dustindall)
- [@yevgnen](https://github.com/Yevgnen)
- [@meldonization](https://github.com/meldonization)
- [Oisín Moran](https://github.com/OisinMoran)
- [Samkit Jain](https://github.com/samkit-jain)
- [Francisco Aranda](https://github.com/frascuchon)
- [Kwok-kuen Cheung](https://github.com/cheungpat)
- [Marco](https://github.com/ubmarco)
- [Idan David](https://github.com/idan-david)
- [@xv44586](https://github.com/xv44586)
- [Alexander Regueiro](https://github.com/alexreg)
- [Daniel Peña](https://github.com/trifling)
- [@bobluda](https://github.com/bobluda)
- [@ramcdona](https://github.com/ramcdona)
- [@johnhuge](https://github.com/johnhuge)
- [Jhonatan Lopes](https://github.com/jhonatan-lopes)
- [Ethan Corey](https://github.com/ethanscorey)
- [Shannon Shen](https://github.com/lolipopshock)
- [Matsumoto Toshi](https://github.com/toshi1127)
- [John West](https://github.com/jwestwsj)
- [David Huggins-Daines](https://github.com/dhdaines)
- [Jeremy B. Merrill](https://github.com/jeremybmerrill)
- [Echedey Luis](https://github.com/echedey-ls)
- [Andy Friedman](https://github.com/afriedman412)
- [Aron Weiler](https://github.com/aronweiler)
- [Quentin André](https://github.com/QuentinAndre11)
- [Léo Roux](https://github.com/leorouxx)
- [@wodny](https://github.com/wodny)
- [Michal Stolarczyk](https://github.com/stolarczyk)
- [Brandon Roberts](https://github.com/brandonrobertz)
- [@ennamarie19](https://github.com/ennamarie19)
- [Anton Ilin](https://github.com/bronislav)
## Contributing
Pull requests are welcome, but please submit a proposal issue first, as the library is in active development.
Current maintainers:
- [Jeremy Singer-Vine](https://github.com/jsvine)
- [Samkit Jain](https://github.com/samkit-jain)
================================================
FILE: codecov.yml
================================================
codecov:
branch: stable
================================================
FILE: docs/colors.md
================================================
# Colors
In the PDF specification, as well as in `pdfplumber`, most graphical objects can have two color attributes:
- `stroking_color`: The color of the object's outline
- `non_stroking_color`: The color of the object's interior, or "fill"
In the PDF specification, colors have both a "color space" and a "color value".
## Color Spaces
Valid color spaces are grouped into three categories:
- Device color spaces
- `DeviceGray`
- `DeviceRGB`
- `DeviceCMYK`
- CIE-based color spaces
- `CalGray`
- `CalRGB`
- `Lab`
- `ICCBased`
- Special color spaces
- `Indexed`
- `Pattern`
- `Separation`
- `DeviceN`
To read more about the differences between those color spaces, see section 4.5 [here](https://ghostscript.com/~robin/pdf_reference17.pdf).
`pdfplumber` aims to expose those color spaces as `scs` (stroking color space) and `ncs` (non-stroking color space), represented as a __string__.
__Caveat__: The only information `pdfplumber` can __currently__ expose is the non-stroking color space for `char` objects. The rest (stroking color space for `char` objects and either color space for the other types of objects) will require a pull request to `pdfminer.six`.
## Color Values
The color value determines *what specific color* in the color space should be used. With the exception of the "special color spaces," these color values are specified as a series of numbers. For `DeviceRGB`, for example, the color values are three numbers, representing the intensities of red, green, and blue.
In `pdfplumber`, those color values are exposed as `stroking_color` and `non_stroking_color`, represented as a __tuple of numbers__.
The pattern specified by the `Pattern` color space is exposed via the `non_stroking_pattern` and `stroking_pattern` attributes.
================================================
FILE: docs/repairing.md
================================================
# Repairing Malformed PDFs
Many parsing issues can be traced back to malformed PDFs.
Malformed PDFs can often be [fixed via Ghostscript](https://superuser.com/questions/278562/how-can-i-fix-repair-a-corrupted-pdf-file).
`pdfplumber` lets you automatically run those repairs, in several ways:
- `pdfplumber.open(..., repair=True)` will repair your PDF on the fly (but not save the repaired version to disk).
- `pdfplumber.repair(path_to_pdf)` will return a `BytesIO` object holding the bytes of a repaired version of the original file.
- `pdfplumber.repair(path_to_pdf, outfile="path/to/repaired.pdf")` will write a repaired version of the original file to the indicated `outfile` path.
## Custom parameters
- `gs_path=...`: You can pass a custom path for the Ghostscript executable, helpful in case `pdfplumber` is unable to auto-detect your copy of Ghostscript.
================================================
FILE: docs/structure.md
================================================
# Structure Tree
Since PDF 1.3 it is possible for a PDF to contain logical structure,
contained in a *structure tree*. In conjunction with PDF 1.2 [marked
content sections](#marked-content-sections) this forms the basis of
Tagged PDF and other accessibility features.
Unfortunately, since all of these standards are optional and variably
implemented in PDF authoring tools, and are frequently not enabled by
default, it is not possible to rely on them to extract the structure
of a PDF and associated content. Nonetheless they can be useful as
features for a heuristic or machine-learning based system, or for
extracting particular structures such as tables.
Since `pdfplumber`'s API is page-based, the structure is available for
a particular page, using the `structure_tree` attribute:
with pdfplumber.open(pdffile) as pdf:
for element in pdf.pages[0].structure_tree:
print(element["type"], element["mcids"])
for child in element.children:
print(child["type"], child["mcids"])
The `type` field contains the type of the structure element - the
standard structure types can be seen in section 10.7.3 of [the PDF 1.7
reference
document](https://ghostscript.com/~robin/pdf_reference17.pdf#page=898),
but usually they are rather HTML-like, if created by a recent PDF
authoring tool (notably, older tools may simply produce `P` for
everything).
The `mcids` field contains the list of marked content section IDs
corresponding to this element.
The `lang` field is often present as well, and contains a language
code for the text content, e.g. `"EN-US"` or `"FR-CA"`.
The `alt_text` field will be present if the author has helpfully added
alternate text to an image. In some cases, `actual_text` may also be
present.
There are also various attributes that may be in the `attributes`
field. Some of these are quite useful indeed, such as ``BBox` which
gives you the bounding box of a `Table`, `Figure`, or `Image`. You
can see a full list of these [in the PDF
spec](https://ghostscript.com/~robin/pdf_reference17.pdf#page=916).
Note that the `BBox` is in PDF coordinate space with the origin at the
bottom left of the page. To convert it to `pdfplumber`'s space you
can do, for example:
x0, y0, x1, y1 = element['attributes']['BBox']
top = page.height - y1
bottom = page.height - y0
doctop = page.initial_doctop + top
bbox = (x0, top, x1, bottom)
It is also possible to get the structure tree for the entire document.
In this case, because marked content IDs are specific to a given page,
each element will also have a `page_number` attribute, which is the
number of the page containing (partially or completely) this element,
indexed from 1 (for consistency with `pdfplumber.Page`).
You can also access the underlying `PDFStructTree` object for more
flexibility, including visual debugging. For instance to plot the
bounding boxes of the contents of all of the `TD` elements on the
first page of a document:
page = pdf.pages[0]
stree = PDFStructTree(pdf, page)
img = page.to_image()
img.draw_rects(stree.element_bbox(td) for td in table.find_all("TD"))
The `find_all` method works rather like the same method in
[BeautifulSoup](https://beautiful-soup-4.readthedocs.io/en/latest/#searching-the-tree) -
it takes an element name, a regular expression, or a matching
function.
================================================
FILE: pdfplumber/__init__.py
================================================
__all__ = [
"__version__",
"utils",
"pdfminer",
"open",
"repair",
"set_debug",
]
import pdfminer
import pdfminer.pdftypes
from . import utils
from ._version import __version__
from .pdf import PDF
from .repair import repair
open = PDF.open
================================================
FILE: pdfplumber/_typing.py
================================================
from typing import Any, Dict, Iterable, List, Literal, Sequence, Tuple, Union
T_seq = Sequence
T_num = Union[int, float]
T_point = Tuple[T_num, T_num]
T_bbox = Tuple[T_num, T_num, T_num, T_num]
T_obj = Dict[str, Any]
T_obj_list = List[T_obj]
T_obj_iter = Iterable[T_obj]
T_dir = Union[Literal["ltr"], Literal["rtl"], Literal["ttb"], Literal["btt"]]
================================================
FILE: pdfplumber/_version.py
================================================
version_info = (0, 11, 9)
__version__ = ".".join(map(str, version_info))
================================================
FILE: pdfplumber/cli.py
================================================
#!/usr/bin/env python
import argparse
import json
import sys
from collections import defaultdict, deque
from itertools import chain
from typing import Any, DefaultDict, Dict, List
from .pdf import PDF
if len(sys.argv) == 1:
sys.argv.append("--help")
def parse_page_spec(p_str: str) -> List[int]:
if "-" in p_str:
start, end = map(int, p_str.split("-"))
return list(range(start, end + 1))
else:
return [int(p_str)]
def parse_args(args_raw: List[str]) -> argparse.Namespace:
parser = argparse.ArgumentParser("pdfplumber")
parser.add_argument("infile", nargs="?", type=argparse.FileType("rb"))
group = parser.add_mutually_exclusive_group()
group.add_argument(
"--structure",
help="Write the structure tree as JSON. "
"All other arguments except --pages, --laparams, and --indent will be ignored",
action="store_true",
)
group.add_argument(
"--structure-text",
help="Write the structure tree as JSON including text contents. "
"All other arguments except --pages, --laparams, and --indent will be ignored",
action="store_true",
)
parser.add_argument("--format", choices=["csv", "json", "text"], default="csv")
parser.add_argument("--types", nargs="+")
parser.add_argument(
"--include-attrs",
nargs="+",
help="Include *only* these object attributes in output.",
)
parser.add_argument(
"--exclude-attrs",
nargs="+",
help="Exclude these object attributes from output.",
)
parser.add_argument("--laparams", type=json.loads)
parser.add_argument("--precision", type=int)
parser.add_argument("--pages", nargs="+", type=parse_page_spec)
parser.add_argument(
"--indent", type=int, help="Indent level for JSON pretty-printing."
)
args = parser.parse_args(args_raw)
if args.pages is not None:
args.pages = list(chain(*args.pages))
return args
def add_text_to_mcids(pdf: PDF, data: List[Dict[str, Any]]) -> None:
page_contents: DefaultDict[int, Any] = defaultdict(lambda: defaultdict(str))
for page in pdf.pages:
text_contents = page_contents[page.page_number]
for c in page.chars:
mcid = c.get("mcid")
if mcid is None:
continue
text_contents[mcid] += c["text"]
d = deque(data)
while d:
el = d.popleft()
if "children" in el:
d.extend(el["children"])
pageno = el.get("page_number")
if pageno is None:
continue
text_contents = page_contents[pageno]
if "mcids" in el:
el["text"] = [text_contents[mcid] for mcid in el["mcids"]]
def main(args_raw: List[str] = sys.argv[1:]) -> None:
args = parse_args(args_raw)
with PDF.open(args.infile, pages=args.pages, laparams=args.laparams) as pdf:
if args.structure:
print(json.dumps(pdf.structure_tree, indent=args.indent))
elif args.structure_text:
tree = pdf.structure_tree
add_text_to_mcids(pdf, tree)
print(json.dumps(tree, indent=args.indent, ensure_ascii=False))
elif args.format == "csv":
pdf.to_csv(
sys.stdout,
args.types,
precision=args.precision,
include_attrs=args.include_attrs,
exclude_attrs=args.exclude_attrs,
)
elif args.format == "text":
for page in pdf.pages:
print(page.extract_text(layout=True))
else:
pdf.to_json(
sys.stdout,
args.types,
precision=args.precision,
include_attrs=args.include_attrs,
exclude_attrs=args.exclude_attrs,
indent=args.indent,
)
if __name__ == "__main__":
main()
================================================
FILE: pdfplumber/container.py
================================================
import csv
import json
from io import StringIO
from itertools import chain
from typing import Any, Dict, List, Optional, Set, TextIO
from . import utils
from ._typing import T_obj, T_obj_list
from .convert import CSV_COLS_REQUIRED, CSV_COLS_TO_PREPEND, Serializer
class Container(object):
cached_properties = ["_rect_edges", "_curve_edges", "_edges", "_objects"]
@property
def pages(self) -> Optional[List[Any]]: # pragma: nocover
raise NotImplementedError
@property
def objects(self) -> Dict[str, T_obj_list]: # pragma: nocover
raise NotImplementedError
def to_dict(
self, object_types: Optional[List[str]] = None
) -> Dict[str, Any]: # pragma: nocover
raise NotImplementedError
def flush_cache(self, properties: Optional[List[str]] = None) -> None:
props = self.cached_properties if properties is None else properties
for p in props:
if hasattr(self, p):
delattr(self, p)
@property
def rects(self) -> T_obj_list:
return self.objects.get("rect", [])
@property
def lines(self) -> T_obj_list:
return self.objects.get("line", [])
@property
def curves(self) -> T_obj_list:
return self.objects.get("curve", [])
@property
def images(self) -> T_obj_list:
return self.objects.get("image", [])
@property
def chars(self) -> T_obj_list:
return self.objects.get("char", [])
@property
def textboxverticals(self) -> T_obj_list:
return self.objects.get("textboxvertical", [])
@property
def textboxhorizontals(self) -> T_obj_list:
return self.objects.get("textboxhorizontal", [])
@property
def textlineverticals(self) -> T_obj_list:
return self.objects.get("textlinevertical", [])
@property
def textlinehorizontals(self) -> T_obj_list:
return self.objects.get("textlinehorizontal", [])
@property
def rect_edges(self) -> T_obj_list:
if hasattr(self, "_rect_edges"):
return self._rect_edges
rect_edges_gen = (utils.rect_to_edges(r) for r in self.rects)
self._rect_edges: T_obj_list = list(chain(*rect_edges_gen))
return self._rect_edges
@property
def curve_edges(self) -> T_obj_list:
if hasattr(self, "_curve_edges"):
return self._curve_edges
curve_edges_gen = (utils.curve_to_edges(r) for r in self.curves)
self._curve_edges: T_obj_list = list(chain(*curve_edges_gen))
return self._curve_edges
@property
def edges(self) -> T_obj_list:
if hasattr(self, "_edges"):
return self._edges
line_edges = list(map(utils.line_to_edge, self.lines))
self._edges: T_obj_list = line_edges + self.rect_edges + self.curve_edges
return self._edges
@property
def horizontal_edges(self) -> T_obj_list:
def test(x: T_obj) -> bool:
return bool(x["orientation"] == "h")
return list(filter(test, self.edges))
@property
def vertical_edges(self) -> T_obj_list:
def test(x: T_obj) -> bool:
return bool(x["orientation"] == "v")
return list(filter(test, self.edges))
def to_json(
self,
stream: Optional[TextIO] = None,
object_types: Optional[List[str]] = None,
include_attrs: Optional[List[str]] = None,
exclude_attrs: Optional[List[str]] = None,
precision: Optional[int] = None,
indent: Optional[int] = None,
) -> Optional[str]:
data = self.to_dict(object_types)
serialized = Serializer(
precision=precision,
include_attrs=include_attrs,
exclude_attrs=exclude_attrs,
).serialize(data)
if stream is None:
return json.dumps(serialized, indent=indent)
else:
json.dump(serialized, stream, indent=indent)
return None
def to_csv(
self,
stream: Optional[TextIO] = None,
object_types: Optional[List[str]] = None,
precision: Optional[int] = None,
include_attrs: Optional[List[str]] = None,
exclude_attrs: Optional[List[str]] = None,
) -> Optional[str]:
if stream is None:
stream = StringIO()
to_string = True
else:
to_string = False
if object_types is None:
object_types = list(self.objects.keys()) + ["annot"]
serialized = []
fields: Set[str] = set()
pages = [self] if self.pages is None else self.pages
serializer = Serializer(
precision=precision,
include_attrs=include_attrs,
exclude_attrs=exclude_attrs,
)
for page in pages:
for t in object_types:
objs = getattr(page, t + "s")
if len(objs):
serialized += serializer.serialize(objs)
new_keys = [k for k, v in objs[0].items() if type(v) is not dict]
fields = fields.union(set(new_keys))
non_req_cols = CSV_COLS_TO_PREPEND + list(
sorted(set(fields) - set(CSV_COLS_REQUIRED + CSV_COLS_TO_PREPEND))
)
cols = CSV_COLS_REQUIRED + list(filter(serializer.attr_filter, non_req_cols))
w = csv.DictWriter(
stream,
fieldnames=cols,
extrasaction="ignore",
quoting=csv.QUOTE_MINIMAL,
escapechar="\\",
)
w.writeheader()
w.writerows(serialized)
if to_string:
stream.seek(0)
return stream.read()
else:
return None
================================================
FILE: pdfplumber/convert.py
================================================
import base64
from typing import Any, Callable, Dict, List, Optional, Tuple
from pdfminer.psparser import PSLiteral
from .utils import decode_text
ENCODINGS_TO_TRY = [
"utf-8",
"latin-1",
"utf-16",
"utf-16le",
]
CSV_COLS_REQUIRED = [
"object_type",
]
CSV_COLS_TO_PREPEND = [
"page_number",
"x0",
"x1",
"y0",
"y1",
"doctop",
"top",
"bottom",
"width",
"height",
]
def get_attr_filter(
include_attrs: Optional[List[str]] = None, exclude_attrs: Optional[List[str]] = None
) -> Callable[[str], bool]:
if include_attrs is not None and exclude_attrs is not None:
raise ValueError(
"Cannot specify `include_attrs` and `exclude_attrs` at the same time."
)
elif include_attrs is not None:
incl = set(CSV_COLS_REQUIRED + include_attrs)
return lambda attr: attr in incl
elif exclude_attrs is not None:
nonexcludable = set(exclude_attrs).intersection(set(CSV_COLS_REQUIRED))
if len(nonexcludable):
raise ValueError(
f"Cannot exclude these required properties: {list(nonexcludable)}"
)
excl = set(exclude_attrs)
return lambda attr: attr not in excl
else:
return lambda attr: True
def to_b64(data: bytes) -> str:
return base64.b64encode(data).decode("ascii")
class Serializer:
def __init__(
self,
precision: Optional[int] = None,
include_attrs: Optional[List[str]] = None,
exclude_attrs: Optional[List[str]] = None,
):
self.precision = precision
self.attr_filter = get_attr_filter(
include_attrs=include_attrs, exclude_attrs=exclude_attrs
)
def serialize(self, obj: Any) -> Any:
if obj is None:
return None
t = type(obj)
# Basic types don't need to be converted
if t in (int, str):
return obj
# Use one of the custom converters, if possible
fn = getattr(self, f"do_{t.__name__}", None)
if fn is not None:
return fn(obj)
# Otherwise, just use the string-representation
else:
return str(obj)
def do_float(self, x: float) -> float:
return x if self.precision is None else round(x, self.precision)
def do_bool(self, x: bool) -> int:
return int(x)
def do_list(self, obj: List[Any]) -> List[Any]:
return list(self.serialize(x) for x in obj)
def do_tuple(self, obj: Tuple[Any, ...]) -> Tuple[Any, ...]:
return tuple(self.serialize(x) for x in obj)
def do_dict(self, obj: Dict[str, Any]) -> Dict[str, Any]:
if "object_type" in obj.keys():
return {k: self.serialize(v) for k, v in obj.items() if self.attr_filter(k)}
else:
return {k: self.serialize(v) for k, v in obj.items()}
def do_PDFStream(self, obj: Any) -> Dict[str, Optional[str]]:
return {"rawdata": to_b64(obj.rawdata) if obj.rawdata else None}
def do_PSLiteral(self, obj: PSLiteral) -> str:
return decode_text(obj.name)
def do_bytes(self, obj: bytes) -> Optional[str]:
for e in ENCODINGS_TO_TRY:
try:
return obj.decode(e)
except UnicodeDecodeError: # pragma: no cover
return None
# If none of the decodings work, raise whatever error
# decoding with utf-8 causes
obj.decode(ENCODINGS_TO_TRY[0]) # pragma: no cover
return None # pragma: no cover
================================================
FILE: pdfplumber/ctm.py
================================================
import math
from typing import NamedTuple
# For more details, see the PDF Reference, 6th Ed., Section 4.2.2 ("Common
# Transformations")
class CTM(NamedTuple):
a: float
b: float
c: float
d: float
e: float
f: float
@property
def scale_x(self) -> float:
return math.sqrt(pow(self.a, 2) + pow(self.b, 2))
@property
def scale_y(self) -> float:
return math.sqrt(pow(self.c, 2) + pow(self.d, 2))
@property
def skew_x(self) -> float:
return (math.atan2(self.d, self.c) * 180 / math.pi) - 90
@property
def skew_y(self) -> float:
return math.atan2(self.b, self.a) * 180 / math.pi
@property
def translation_x(self) -> float:
return self.e
@property
def translation_y(self) -> float:
return self.f
================================================
FILE: pdfplumber/display.py
================================================
import pathlib
from io import BufferedReader, BytesIO
from typing import TYPE_CHECKING, Any, List, Optional, Tuple, Union
import PIL.Image
import PIL.ImageDraw
import pypdfium2 # type: ignore
from . import utils
from ._typing import T_bbox, T_num, T_obj, T_obj_list, T_point, T_seq
from .table import T_table_settings, Table, TableFinder, TableSettings
from .utils.exceptions import MalformedPDFException
if TYPE_CHECKING: # pragma: nocover
import pandas as pd
from .page import Page
class COLORS:
RED = (255, 0, 0)
GREEN = (0, 255, 0)
BLUE = (0, 0, 255)
TRANSPARENT = (0, 0, 0, 0)
DEFAULT_FILL = COLORS.BLUE + (50,)
DEFAULT_STROKE = COLORS.RED + (200,)
DEFAULT_STROKE_WIDTH = 1
DEFAULT_RESOLUTION = 72
T_color = Union[Tuple[int, int, int], Tuple[int, int, int, int], str]
T_contains_points = Union[Tuple[T_point, ...], List[T_point], T_obj]
def get_page_image(
stream: Union[BufferedReader, BytesIO],
path: Optional[pathlib.Path],
page_ix: int,
resolution: Union[int, float],
password: Optional[str],
antialias: bool = False,
) -> PIL.Image.Image:
src: Union[pathlib.Path, BufferedReader, BytesIO]
# If we are working with a file object saved to disk
if path:
src = path
# If we instead are working with a BytesIO stream
else:
stream.seek(0)
src = stream
try:
pdfium_doc = pypdfium2.PdfDocument(src, password=password)
except pypdfium2.PdfiumError as e:
raise MalformedPDFException(e)
pdfium_page = pdfium_doc.get_page(page_ix)
img: PIL.Image.Image = pdfium_page.render(
# Modifiable arguments
scale=resolution / 72,
no_smoothtext=not antialias,
no_smoothpath=not antialias,
no_smoothimage=not antialias,
# Non-modifiable arguments
prefer_bgrx=True,
).to_pil()
pdfium_doc.close()
return img.convert("RGB")
class PageImage:
def __init__(
self,
page: "Page",
original: Optional[PIL.Image.Image] = None,
resolution: Union[int, float] = DEFAULT_RESOLUTION,
antialias: bool = False,
force_mediabox: bool = False,
):
self.page = page
self.root = page if page.is_original else page.root_page
self.resolution = resolution
if original is None:
self.original = get_page_image(
stream=page.pdf.stream,
path=page.pdf.path,
page_ix=page.page_number - 1,
resolution=resolution,
antialias=antialias,
password=page.pdf.password,
)
else:
self.original = original
self.scale = self.original.size[0] / (page.cropbox[2] - page.cropbox[0])
# This value represents the coordinates of the page,
# in page-unit values, that will be displayed.
self.bbox = (
page.bbox
if page.bbox != page.mediabox
else (page.mediabox if force_mediabox else page.cropbox)
)
# If this value is different than the *Page*'s .cropbox
# (e.g., because the mediabox differs from the cropbox or
# or because we've used Page.crop(...)), then we'll need to
# crop the initially-converted image.
if page.bbox != page.cropbox:
crop_dims = self._reproject_bbox(page.cropbox)
bbox_dims = self._reproject_bbox(self.bbox)
self.original = self.original.crop(
(
bbox_dims[0] - crop_dims[0],
bbox_dims[1] - crop_dims[1],
bbox_dims[2] - crop_dims[0],
bbox_dims[3] - crop_dims[1],
)
)
self.reset()
def _reproject_bbox(self, bbox: T_bbox) -> Tuple[int, int, int, int]:
x0, top, x1, bottom = bbox
_x0, _top = self._reproject((x0, top))
_x1, _bottom = self._reproject((x1, bottom))
return (_x0, _top, _x1, _bottom)
def _reproject(self, coord: T_point) -> Tuple[int, int]:
"""
Given an (x0, top) tuple from the *root* coordinate system,
return an (x0, top) tuple in the *image* coordinate system.
"""
x0, top = coord
_x0 = (x0 - self.bbox[0]) * self.scale
_top = (top - self.bbox[1]) * self.scale
return (int(_x0), int(_top))
def reset(self) -> "PageImage":
self.annotated = PIL.Image.new("RGB", self.original.size)
self.annotated.paste(self.original)
self.draw = PIL.ImageDraw.Draw(self.annotated, "RGBA")
return self
def save(
self,
dest: Union[str, pathlib.Path, BytesIO],
format: str = "PNG",
quantize: bool = True,
colors: int = 256,
bits: int = 8,
**kwargs: Any,
) -> None:
if quantize:
out = self.annotated.quantize(colors, method=PIL.Image.FASTOCTREE).convert(
"P"
)
else:
out = self.annotated
out.save(
dest,
format=format,
bits=bits,
dpi=(self.resolution, self.resolution),
**kwargs,
)
def copy(self) -> "PageImage":
return self.__class__(self.page, self.original)
def draw_line(
self,
points_or_obj: T_contains_points,
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
# If passing a raw list of points, use those
if isinstance(points_or_obj, (tuple, list)):
points = points_or_obj
# Else, use the "pts" attribute if available
elif isinstance(points_or_obj, dict) and "pts" in points_or_obj:
points = [(x, y) for x, y in points_or_obj["pts"]]
# Otherwise, just use ((x0, top), (x1, bottom))
else:
obj = points_or_obj
points = ((obj["x0"], obj["top"]), (obj["x1"], obj["bottom"]))
self.draw.line(
list(map(self._reproject, points)), fill=stroke, width=stroke_width
)
return self
def draw_lines(
self,
list_of_lines: Union[T_seq[T_contains_points], "pd.DataFrame"],
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
for x in utils.to_list(list_of_lines):
self.draw_line(x, stroke=stroke, stroke_width=stroke_width)
return self
def draw_vline(
self,
location: T_num,
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
points = (location, self.bbox[1], location, self.bbox[3])
self.draw.line(self._reproject_bbox(points), fill=stroke, width=stroke_width)
return self
def draw_vlines(
self,
locations: Union[List[T_num], "pd.Series[float]"],
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
for x in list(locations):
self.draw_vline(x, stroke=stroke, stroke_width=stroke_width)
return self
def draw_hline(
self,
location: T_num,
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
points = (self.bbox[0], location, self.bbox[2], location)
self.draw.line(self._reproject_bbox(points), fill=stroke, width=stroke_width)
return self
def draw_hlines(
self,
locations: Union[List[T_num], "pd.Series[float]"],
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
for x in list(locations):
self.draw_hline(x, stroke=stroke, stroke_width=stroke_width)
return self
def draw_rect(
self,
bbox_or_obj: Union[T_bbox, T_obj],
fill: T_color = DEFAULT_FILL,
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
if isinstance(bbox_or_obj, (tuple, list)):
bbox = bbox_or_obj
else:
obj = bbox_or_obj
bbox = (obj["x0"], obj["top"], obj["x1"], obj["bottom"])
x0, top, x1, bottom = bbox
half = stroke_width / 2
x0 = min(x0 + half, (x0 + x1) / 2)
top = min(top + half, (top + bottom) / 2)
x1 = max(x1 - half, (x0 + x1) / 2)
bottom = max(bottom - half, (top + bottom) / 2)
fill_bbox = self._reproject_bbox((x0, top, x1, bottom))
self.draw.rectangle(fill_bbox, fill, COLORS.TRANSPARENT)
if stroke_width > 0:
segments = [
((x0, top), (x1, top)), # top
((x0, bottom), (x1, bottom)), # bottom
((x0, top), (x0, bottom)), # left
((x1, top), (x1, bottom)), # right
]
self.draw_lines(segments, stroke=stroke, stroke_width=stroke_width)
return self
def draw_rects(
self,
list_of_rects: Union[List[T_bbox], T_obj_list, "pd.DataFrame"],
fill: T_color = DEFAULT_FILL,
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
for x in utils.to_list(list_of_rects):
self.draw_rect(x, fill=fill, stroke=stroke, stroke_width=stroke_width)
return self
def draw_circle(
self,
center_or_obj: Union[T_point, T_obj],
radius: int = 5,
fill: T_color = DEFAULT_FILL,
stroke: T_color = DEFAULT_STROKE,
) -> "PageImage":
if isinstance(center_or_obj, tuple):
center = center_or_obj
else:
obj = center_or_obj
center = ((obj["x0"] + obj["x1"]) / 2, (obj["top"] + obj["bottom"]) / 2)
cx, cy = center
bbox = (cx - radius, cy - radius, cx + radius, cy + radius)
self.draw.ellipse(self._reproject_bbox(bbox), fill, stroke)
return self
def draw_circles(
self,
list_of_circles: Union[List[T_point], T_obj_list, "pd.DataFrame"],
radius: int = 5,
fill: T_color = DEFAULT_FILL,
stroke: T_color = DEFAULT_STROKE,
) -> "PageImage":
for x in utils.to_list(list_of_circles):
self.draw_circle(x, radius=radius, fill=fill, stroke=stroke)
return self
def debug_table(
self,
table: Table,
fill: T_color = DEFAULT_FILL,
stroke: T_color = DEFAULT_STROKE,
stroke_width: int = 1,
) -> "PageImage":
"""
Outline all found tables.
"""
self.draw_rects(
table.cells, fill=fill, stroke=stroke, stroke_width=stroke_width
)
return self
def debug_tablefinder(
self,
table_settings: Optional[
Union[TableFinder, TableSettings, T_table_settings]
] = None,
) -> "PageImage":
if isinstance(table_settings, TableFinder):
finder = table_settings
elif table_settings is None or isinstance(
table_settings, (TableSettings, dict)
):
finder = self.page.debug_tablefinder(table_settings)
else:
raise ValueError(
"Argument must be instance of TableFinder"
"or a TableFinder settings dict."
)
for table in finder.tables:
self.debug_table(table)
self.draw_lines(finder.edges, stroke_width=1)
self.draw_circles(
list(finder.intersections.keys()),
fill=COLORS.TRANSPARENT,
stroke=COLORS.BLUE + (200,),
radius=3,
)
return self
def outline_words(
self,
stroke: T_color = DEFAULT_STROKE,
fill: T_color = DEFAULT_FILL,
stroke_width: int = DEFAULT_STROKE_WIDTH,
x_tolerance: T_num = utils.DEFAULT_X_TOLERANCE,
y_tolerance: T_num = utils.DEFAULT_Y_TOLERANCE,
) -> "PageImage":
words = self.page.extract_words(
x_tolerance=x_tolerance, y_tolerance=y_tolerance
)
self.draw_rects(words, stroke=stroke, fill=fill, stroke_width=stroke_width)
return self
def outline_chars(
self,
stroke: T_color = (255, 0, 0, 255),
fill: T_color = (255, 0, 0, int(255 / 4)),
stroke_width: int = DEFAULT_STROKE_WIDTH,
) -> "PageImage":
self.draw_rects(
self.page.chars, stroke=stroke, fill=fill, stroke_width=stroke_width
)
return self
def _repr_png_(self) -> bytes:
b = BytesIO()
self.save(b, "PNG")
return b.getvalue()
def show(self) -> None: # pragma: no cover
self.annotated.show()
================================================
FILE: pdfplumber/page.py
================================================
import numbers
import re
from functools import lru_cache
from typing import (
TYPE_CHECKING,
Any,
Callable,
Dict,
Generator,
List,
Optional,
Pattern,
Tuple,
Union,
)
from unicodedata import normalize as normalize_unicode
from warnings import warn
from pdfminer.converter import PDFPageAggregator
from pdfminer.layout import (
LTChar,
LTComponent,
LTContainer,
LTCurve,
LTItem,
LTPage,
LTTextContainer,
)
from pdfminer.pdfinterp import PDFPageInterpreter, PDFStackT
from pdfminer.pdfpage import PDFPage
from pdfminer.psparser import PSLiteral
from . import utils
from ._typing import T_bbox, T_num, T_obj, T_obj_list
from .container import Container
from .structure import PDFStructTree, StructTreeMissing
from .table import T_table_settings, Table, TableFinder, TableSettings
from .utils import decode_text, resolve_all, resolve_and_decode
from .utils.exceptions import MalformedPDFException, PdfminerException
from .utils.text import TextMap
lt_pat = re.compile(r"^LT")
ALL_ATTRS = set(
[
"adv",
"height",
"linewidth",
"pts",
"size",
"srcsize",
"width",
"x0",
"x1",
"y0",
"y1",
"bits",
"matrix",
"upright",
"fontname",
"text",
"imagemask",
"colorspace",
"evenodd",
"fill",
"non_stroking_color",
"stroke",
"stroking_color",
"stream",
"name",
"mcid",
"tag",
]
)
if TYPE_CHECKING: # pragma: nocover
from .display import PageImage
from .pdf import PDF
# via https://git.ghostscript.com/?p=mupdf.git;a=blob;f=source/pdf/pdf-font.c;h=6322cedf2c26cfb312c0c0878d7aff97b4c7470e;hb=HEAD#l774 # noqa
CP936_FONTNAMES = {
b"\xcb\xce\xcc\xe5": "SimSun,Regular",
b"\xba\xda\xcc\xe5": "SimHei,Regular",
b"\xbf\xac\xcc\xe5_GB2312": "SimKai,Regular",
b"\xb7\xc2\xcb\xce_GB2312": "SimFang,Regular",
b"\xc1\xa5\xca\xe9": "SimLi,Regular",
}
def fix_fontname_bytes(fontname: bytes) -> str:
if b"+" in fontname:
split_at = fontname.index(b"+") + 1
prefix, suffix = fontname[:split_at], fontname[split_at:]
else:
prefix, suffix = b"", fontname
suffix_new = CP936_FONTNAMES.get(suffix, str(suffix)[2:-1])
return str(prefix)[2:-1] + suffix_new
def tuplify_list_kwargs(kwargs: Dict[str, Any]) -> Dict[str, Any]:
return {
key: (tuple(value) if isinstance(value, list) else value)
for key, value in kwargs.items()
}
class PDFPageAggregatorWithMarkedContent(PDFPageAggregator):
"""Extract layout from a specific page, adding marked-content IDs to
objects where found."""
cur_mcid: Optional[int] = None
cur_tag: Optional[str] = None
def begin_tag(self, tag: PSLiteral, props: Optional[PDFStackT] = None) -> None:
"""Handle beginning of tag, setting current MCID if any."""
self.cur_tag = decode_text(tag.name)
if isinstance(props, dict) and "MCID" in props:
self.cur_mcid = props["MCID"]
else:
self.cur_mcid = None
def end_tag(self) -> None:
"""Handle beginning of tag, clearing current MCID."""
self.cur_tag = None
self.cur_mcid = None
def tag_cur_item(self) -> None:
"""Add current MCID to what we hope to be the most recent object created
by pdfminer.six."""
# This is somewhat hacky and would not be necessary if
# pdfminer.six supported MCIDs. In reading the code it's
# clear that the `render_*` methods methods will only ever
# create one object, but that is far from being guaranteed.
# Even if pdfminer.six's API would just return the objects it
# creates, we wouldn't have to do this.
if self.cur_item._objs:
cur_obj = self.cur_item._objs[-1]
cur_obj.mcid = self.cur_mcid # type: ignore
cur_obj.tag = self.cur_tag # type: ignore
def render_char(self, *args, **kwargs) -> float: # type: ignore
"""Hook for rendering characters, adding the `mcid` attribute."""
adv = super().render_char(*args, **kwargs)
self.tag_cur_item()
return adv
def render_image(self, *args, **kwargs) -> None: # type: ignore
"""Hook for rendering images, adding the `mcid` attribute."""
super().render_image(*args, **kwargs)
self.tag_cur_item()
def paint_path(self, *args, **kwargs) -> None: # type: ignore
"""Hook for rendering lines and curves, adding the `mcid` attribute."""
super().paint_path(*args, **kwargs)
self.tag_cur_item()
def _normalize_box(box_raw: T_bbox, rotation: T_num = 0) -> T_bbox:
# Per PDF Reference 3.8.4: "Note: Although rectangles are
# conventionally specified by their lower-left and upperright
# corners, it is acceptable to specify any two diagonally opposite
# corners."
if not all(isinstance(x, numbers.Number) for x in box_raw): # pragma: nocover
raise MalformedPDFException(
f"Bounding box contains non-number coordinate(s): {box_raw}"
)
x0, x1 = sorted((box_raw[0], box_raw[2]))
y0, y1 = sorted((box_raw[1], box_raw[3]))
if rotation in [90, 270]:
return (y0, x0, y1, x1)
else:
return (x0, y0, x1, y1)
# PDFs coordinate spaces refer to an origin in the bottom-left of the
# page; pdfplumber flips this vertically, so that the origin is in the
# top-left.
def _invert_box(box_raw: T_bbox, mb_height: T_num) -> T_bbox:
x0, y0, x1, y1 = box_raw
return (x0, mb_height - y1, x1, mb_height - y0)
class Page(Container):
cached_properties: List[str] = Container.cached_properties + ["_layout"]
is_original: bool = True
pages = None
def __init__(
self,
pdf: "PDF",
page_obj: PDFPage,
page_number: int,
initial_doctop: T_num = 0,
):
self.pdf = pdf
self.root_page = self
self.page_obj = page_obj
self.page_number = page_number
self.initial_doctop = initial_doctop
def get_attr(key: str, default: Any = None) -> Any:
value = resolve_all(page_obj.attrs.get(key))
return default if value is None else value
# Per PDF Reference Table 3.27: "The number of degrees by which the
# page should be rotated clockwise when displayed or printed. The value
# must be a multiple of 90. Default value: 0"
_rotation = get_attr("Rotate", 0)
self.rotation = _rotation % 360
mb_raw = _normalize_box(get_attr("MediaBox"), self.rotation)
mb_height = mb_raw[3] - mb_raw[1]
self.mediabox = _invert_box(mb_raw, mb_height)
for box_name in ["CropBox", "TrimBox", "BleedBox", "ArtBox"]:
if box_name in page_obj.attrs:
box_normalized = _invert_box(
_normalize_box(get_attr(box_name), self.rotation), mb_height
)
setattr(self, box_name.lower(), box_normalized)
if "CropBox" not in page_obj.attrs:
self.cropbox = self.mediabox
# Page.bbox defaults to self.mediabox, but can be altered by Page.crop(...)
self.bbox = self.mediabox
# See https://rednafi.com/python/lru_cache_on_methods/
self.get_textmap = lru_cache()(self._get_textmap)
def close(self) -> None:
self.flush_cache()
self.get_textmap.cache_clear()
@property
def width(self) -> T_num:
return self.bbox[2] - self.bbox[0]
@property
def height(self) -> T_num:
return self.bbox[3] - self.bbox[1]
@property
def structure_tree(self) -> List[Dict[str, Any]]:
"""Return the structure tree for a page, if any."""
try:
return [elem.to_dict() for elem in PDFStructTree(self.pdf, self)]
except StructTreeMissing:
return []
@property
def layout(self) -> LTPage:
if hasattr(self, "_layout"):
return self._layout
device = PDFPageAggregatorWithMarkedContent(
self.pdf.rsrcmgr,
pageno=self.page_number,
laparams=self.pdf.laparams,
)
interpreter = PDFPageInterpreter(self.pdf.rsrcmgr, device)
try:
interpreter.process_page(self.page_obj)
except Exception as e:
raise PdfminerException(e)
self._layout: LTPage = device.get_result()
return self._layout
@property
def annots(self) -> T_obj_list:
def rotate_point(pt: Tuple[float, float], r: int) -> Tuple[float, float]:
turns = r // 90
for i in range(turns):
x, y = pt
comp = self.width if i == turns % 2 else self.height
pt = (y, (comp - x))
return pt
def parse(annot: T_obj) -> T_obj:
_a, _b, _c, _d = annot["Rect"]
pt0 = rotate_point((_a, _b), self.rotation)
pt1 = rotate_point((_c, _d), self.rotation)
rh = self.root_page.height
x0, top, x1, bottom = _invert_box(_normalize_box((*pt0, *pt1)), rh)
a = annot.get("A", {})
extras = {
"uri": a.get("URI"),
"title": annot.get("T"),
"contents": annot.get("Contents"),
}
for k, v in extras.items():
if v is not None:
try:
extras[k] = v.decode("utf-8")
except UnicodeDecodeError:
try:
extras[k] = v.decode("utf-16")
except UnicodeDecodeError:
if self.pdf.raise_unicode_errors:
raise
warn(
f"Could not decode {k} of annotation."
f" {k} will be missing."
)
parsed = {
"page_number": self.page_number,
"object_type": "annot",
"x0": x0,
"y0": rh - bottom,
"x1": x1,
"y1": rh - top,
"doctop": self.initial_doctop + top,
"top": top,
"bottom": bottom,
"width": x1 - x0,
"height": bottom - top,
}
parsed.update(extras)
# Replace the indirect reference to the page dictionary
# with a pointer to our actual page
if "P" in annot:
annot["P"] = self
parsed["data"] = annot
return parsed
raw = resolve_all(self.page_obj.annots) or []
parsed = list(map(parse, raw))
if isinstance(self, CroppedPage):
return self._crop_fn(parsed)
else:
return parsed
@property
def hyperlinks(self) -> T_obj_list:
return [a for a in self.annots if a["uri"] is not None]
@property
def objects(self) -> Dict[str, T_obj_list]:
if hasattr(self, "_objects"):
return self._objects
self._objects: Dict[str, T_obj_list] = self.parse_objects()
return self._objects
def point2coord(self, pt: Tuple[T_num, T_num]) -> Tuple[T_num, T_num]:
# See note below re. #1181 and mediabox-adjustment reversions
return (self.mediabox[0] + pt[0], self.mediabox[1] + self.height - pt[1])
def process_object(self, obj: LTItem) -> T_obj:
kind = re.sub(lt_pat, "", obj.__class__.__name__).lower()
def process_attr(item: Tuple[str, Any]) -> Optional[Tuple[str, Any]]:
k, v = item
if k in ALL_ATTRS:
res = resolve_all(v)
return (k, res)
else:
return None
attr = dict(filter(None, map(process_attr, obj.__dict__.items())))
attr["object_type"] = kind
attr["page_number"] = self.page_number
for cs in ["ncs", "scs"]:
# Note: As of pdfminer.six v20221105, that library only
# exposes ncs for LTChars, and neither attribute for
# other objects. Keeping this code here, though,
# for ease of addition if color spaces become
# more available via pdfminer.six
if hasattr(obj, cs):
attr[cs] = resolve_and_decode(getattr(obj, cs).name)
if isinstance(obj, (LTChar, LTTextContainer)):
text = obj.get_text()
attr["text"] = (
normalize_unicode(self.pdf.unicode_norm, text)
if self.pdf.unicode_norm is not None
else text
)
if isinstance(obj, LTChar):
# pdfminer.six (at least as of v20221105) does not
# directly expose .stroking_color and .non_stroking_color
# for LTChar objects (unlike, e.g., LTRect objects).
gs = obj.graphicstate
attr["stroking_color"] = (
gs.scolor if isinstance(gs.scolor, tuple) else (gs.scolor,)
)
attr["non_stroking_color"] = (
gs.ncolor if isinstance(gs.ncolor, tuple) else (gs.ncolor,)
)
# Handle (rare) byte-encoded fontnames
if isinstance(attr["fontname"], bytes): # pragma: nocover
attr["fontname"] = fix_fontname_bytes(attr["fontname"])
elif isinstance(obj, (LTCurve,)):
attr["pts"] = list(map(self.point2coord, attr["pts"]))
# Ignoring typing because type signature for obj.original_path
# appears to be incorrect
attr["path"] = [(cmd, *map(self.point2coord, pts)) for cmd, *pts in obj.original_path] # type: ignore # noqa: E501
attr["dash"] = obj.dashing_style
# As noted in #1181, `pdfminer.six` adjusts objects'
# coordinates relative to the MediaBox:
# https://github.com/pdfminer/pdfminer.six/blob/1a8bd2f730295b31d6165e4d95fcb5a03793c978/pdfminer/converter.py#L79-L84
mb_x0, mb_top = self.mediabox[:2]
if "y0" in attr:
attr["top"] = (self.height - attr["y1"]) + mb_top
attr["bottom"] = (self.height - attr["y0"]) + mb_top
attr["doctop"] = self.initial_doctop + attr["top"]
if "x0" in attr and mb_x0 != 0:
attr["x0"] = attr["x0"] + mb_x0
attr["x1"] = attr["x1"] + mb_x0
return attr
def iter_layout_objects(
self, layout_objects: List[LTComponent]
) -> Generator[T_obj, None, None]:
for obj in layout_objects:
# If object is, like LTFigure, a higher-level object ...
if isinstance(obj, LTContainer):
# and LAParams is passed, process the object itself.
if self.pdf.laparams is not None:
yield self.process_object(obj)
# Regardless, iterate through its children
yield from self.iter_layout_objects(obj._objs)
else:
yield self.process_object(obj)
def parse_objects(self) -> Dict[str, T_obj_list]:
objects: Dict[str, T_obj_list] = {}
for obj in self.iter_layout_objects(self.layout._objs):
kind = obj["object_type"]
if kind in ["anno"]:
continue
if objects.get(kind) is None:
objects[kind] = []
objects[kind].append(obj)
return objects
def debug_tablefinder(
self, table_settings: Optional[T_table_settings] = None
) -> TableFinder:
tset = TableSettings.resolve(table_settings)
return TableFinder(self, tset)
def find_tables(
self, table_settings: Optional[T_table_settings] = None
) -> List[Table]:
tset = TableSettings.resolve(table_settings)
return TableFinder(self, tset).tables
def find_table(
self, table_settings: Optional[T_table_settings] = None
) -> Optional[Table]:
tset = TableSettings.resolve(table_settings)
tables = self.find_tables(tset)
if len(tables) == 0:
return None
# Return the largest table, as measured by number of cells.
def sorter(x: Table) -> Tuple[int, T_num, T_num]:
return (-len(x.cells), x.bbox[1], x.bbox[0])
largest = list(sorted(tables, key=sorter))[0]
return largest
def extract_tables(
self, table_settings: Optional[T_table_settings] = None
) -> List[List[List[Optional[str]]]]:
tset = TableSettings.resolve(table_settings)
tables = self.find_tables(tset)
return [table.extract(**(tset.text_settings or {})) for table in tables]
def extract_table(
self, table_settings: Optional[T_table_settings] = None
) -> Optional[List[List[Optional[str]]]]:
tset = TableSettings.resolve(table_settings)
table = self.find_table(tset)
if table is None:
return None
else:
return table.extract(**(tset.text_settings or {}))
def _get_textmap(self, **kwargs: Any) -> TextMap:
defaults: Dict[str, Any] = dict(
layout_bbox=self.bbox,
)
if "layout_width_chars" not in kwargs:
defaults.update({"layout_width": self.width})
if "layout_height_chars" not in kwargs:
defaults.update({"layout_height": self.height})
full_kwargs: Dict[str, Any] = {**defaults, **kwargs}
return utils.chars_to_textmap(self.chars, **full_kwargs)
def search(
self,
pattern: Union[str, Pattern[str]],
regex: bool = True,
case: bool = True,
main_group: int = 0,
return_chars: bool = True,
return_groups: bool = True,
**kwargs: Any,
) -> List[Dict[str, Any]]:
textmap = self.get_textmap(**tuplify_list_kwargs(kwargs))
return textmap.search(
pattern,
regex=regex,
case=case,
main_group=main_group,
return_chars=return_chars,
return_groups=return_groups,
)
def extract_text(self, **kwargs: Any) -> str:
return self.get_textmap(**tuplify_list_kwargs(kwargs)).as_string
def extract_text_simple(self, **kwargs: Any) -> str:
return utils.extract_text_simple(self.chars, **kwargs)
def extract_words(self, **kwargs: Any) -> T_obj_list:
return utils.extract_words(self.chars, **kwargs)
def extract_text_lines(
self, strip: bool = True, return_chars: bool = True, **kwargs: Any
) -> T_obj_list:
return self.get_textmap(**tuplify_list_kwargs(kwargs)).extract_text_lines(
strip=strip, return_chars=return_chars
)
def crop(
self, bbox: T_bbox, relative: bool = False, strict: bool = True
) -> "CroppedPage":
return CroppedPage(self, bbox, relative=relative, strict=strict)
def within_bbox(
self, bbox: T_bbox, relative: bool = False, strict: bool = True
) -> "CroppedPage":
"""
Same as .crop, except only includes objects fully within the bbox
"""
return CroppedPage(
self, bbox, relative=relative, strict=strict, crop_fn=utils.within_bbox
)
def outside_bbox(
self, bbox: T_bbox, relative: bool = False, strict: bool = True
) -> "CroppedPage":
"""
Same as .crop, except only includes objects fully within the bbox
"""
return CroppedPage(
self, bbox, relative=relative, strict=strict, crop_fn=utils.outside_bbox
)
def filter(self, test_function: Callable[[T_obj], bool]) -> "FilteredPage":
return FilteredPage(self, test_function)
def dedupe_chars(self, **kwargs: Any) -> "FilteredPage":
"""
Removes duplicate chars — those sharing the same text and positioning
(within `tolerance`) as other characters in the set. Adjust extra_args
to be more/less restrictive with the properties checked.
"""
p = FilteredPage(self, lambda x: True)
p._objects = {kind: objs for kind, objs in self.objects.items()}
p._objects["char"] = utils.dedupe_chars(self.chars, **kwargs)
return p
def to_image(
self,
resolution: Optional[Union[int, float]] = None,
width: Optional[Union[int, float]] = None,
height: Optional[Union[int, float]] = None,
antialias: bool = False,
force_mediabox: bool = False,
) -> "PageImage":
"""
You can pass a maximum of 1 of the following:
- resolution: The desired number pixels per inch. Defaults to 72.
- width: The desired image width in pixels.
- height: The desired image width in pixels.
"""
from .display import DEFAULT_RESOLUTION, PageImage
num_specs = sum(x is not None for x in [resolution, width, height])
if num_specs > 1:
raise ValueError(
f"Only one of these arguments can be provided: resolution, width, height. You provided {num_specs}" # noqa: E501
)
elif width is not None:
resolution = 72 * width / self.width
elif height is not None:
resolution = 72 * height / self.height
return PageImage(
self,
resolution=resolution or DEFAULT_RESOLUTION,
antialias=antialias,
force_mediabox=force_mediabox,
)
def to_dict(self, object_types: Optional[List[str]] = None) -> Dict[str, Any]:
if object_types is None:
_object_types = list(self.objects.keys()) + ["annot"]
else:
_object_types = object_types
d = {
"page_number": self.page_number,
"initial_doctop": self.initial_doctop,
"rotation": self.rotation,
"cropbox": self.cropbox,
"mediabox": self.mediabox,
"bbox": self.bbox,
"width": self.width,
"height": self.height,
}
for t in _object_types:
d[t + "s"] = getattr(self, t + "s")
return d
def __repr__(self) -> str:
return f"