LunarEclipse
/
pdf-table-extractor
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

nattyfix

This commit is contained in:
LunarEclipse 2024-04-27 23:29:33 +02:00
parent 00828d4c08
commit e526de786e
1 changed files with 41 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
pdf_table_extractor.py
View File

@ -3,8 +3,10 @@ import traceback
from typing import Dict, List, Optional, Sequence, Tuple, TypeAlias
import gi
gi.require_version("Gtk", "3.0")
from gi.repository import GLib, Gio, Gtk, GObject
gi.require_foreign("cairo")
import cairo
@ -14,41 +16,46 @@ TEST_FILENAME = "/home/luna/Documents/Resources/Praca Licencjacka/sources/2018_T
Coords: TypeAlias = Tuple[float, float]
class Selection:
def __init__(self, bounds: Tuple[Coords, Coords], columns: Optional[Sequence[float]] = None):
self.bounds = bounds
self.columns = columns
class Document:
def __init__(self, filename: str):
self.filename = filename
self.document = fitz.Document(filename)
self.selections: Dict[int, List[Selection]] = {}
class State():
pass
class PdfPage(Gtk.DrawingArea):
def __init__(self, page, *args, **kwargs):
super().__init__(*args, **kwargs)
self.page: fitz.Page = page
pix = self.page.get_pixmap(dpi=300) # type: ignore
pix = self.page.get_pixmap(dpi=96) # type: ignore
self.set_size_request(pix.width, pix.height)
print(f"w: {pix.width}, h: {pix.height}")
self.set_app_paintable(True) # type: ignore
self.connect("draw", self.on_draw, {})
def on_draw(self, widget, cr: cairo.Context, data: GObject.GPointer):
def on_draw(self, widget: Gtk.DrawingArea, cr: cairo.Context, data: GObject.GPointer):
# app: Application = widget.get_window().get_application() # type: ignore
width = self.get_allocated_width()
height = self.get_allocated_height()
width = widget.get_allocated_width()
height = widget.get_allocated_height()
print(f"alloc w: {width}, h: {height}")
sctx = self.get_style_context()
sctx = widget.get_style_context()
Gtk.render_background(sctx, cr, 0, 0, width, height)
pix = self.page.get_pixmap(dpi=300) # type: ignore
png = pix.tobytes("png")
ims = cairo.ImageSurface.create_from_png(png)
pix = self.page.get_pixmap(dpi=96, alpha=True) # type: ignore
mv: memoryview = pix.samples_mv
ims = cairo.ImageSurface.create_for_data(mv, cairo.Format.RGB24, pix.width, pix.height)
cr.set_source_surface(ims, 0, 0)
cr.paint()
@ -81,7 +88,6 @@ class MainWindow(Gtk.ApplicationWindow):
# self.pdf_list_box.add(PdfPage())
# @Gtk.Template.Callback()
# def example_button_released_cb(self, widget: Gtk.Button, **kwargs):
# assert self.example_button == widget
@ -127,10 +133,12 @@ class MainWindow(Gtk.ApplicationWindow):
self.header_bar.set_subtitle(document.filename) # type: ignore
for i in document.document.pages(): # type: ignore
row = Gtk.ListBoxRow()
page = PdfPage(i)
box = Gtk.Box()
box.pack_start(page, True, True, 20)
self.pdf_list_box.add(box) # type: ignore
row.add(page)
self.pdf_list_box.add(row) # type: ignore
self.pdf_list_box.show_all()
@Gtk.Template.Callback()
def on_open_button_small_clicked(self, widget, **kwargs):