How to extract more than one tables available in the single page pdf as separate data frame or md or csv?

[thangarajdeivasikamani]

Hello Team,

I have the pdf which has more than one tables in single page.
Example :

First table have 5 columns and 10 rows, 2nd Tables have 10 column,3 rows , 3rd Table 3 columns , 20 rows.

In that case how to extract the each table as individual data frame and combine at page level.

[JorjMcKie]

Create an empty list to be used as a list of data frames. Then, for each table encountered, append table.to_pandas() to that list.
Then look up the pandas documentation to see what you can do with separate data frames.

[JorjMcKie]

For example, pandas offers joining data frames, or outputting them to CSV, Excel or also markdown.

[thangarajdeivasikamani]

if my pdf has single Table the below code able to find the Table and creating the data frame.

_[tabs = page.find_tables() # detect the tables
print(tabs)
for i,tab in enumerate(tabs): # iterate over all tables
print(i,tab)
for cell in tab.header.cells:
print(cell)
page.draw_rect(cell,color=fitz.pdfcolor["red"],width=0.3)
print("--")
page.draw_rect(tab.bbox,color=fitz.pdfcolor["green"])
print(f"Table {i} column names: {tab.header.names}, external: {tab.header.external}")

show_image(page, f"Table & Header BBoxes")]_

output

pymupdf.table.TableFinder object at 0x000002A8F9B09590
0 pymupdf.table.Table object at 0x000002A8EA7C2810
(308.22000885009766, 631.5, 380.85880317962426, 649.6500244140625)

(380.85880317962426, 631.5, 528.3300018310547, 649.6500244140625)

Table 0 column names: ['Reference', 'Part number'], external: False

But the pdf page has multiple Tables the same code giving the below error. Not sure what could be wrong.

output

pymupdf.table.TableFinder object at 0x000002A8FADDF410
0 pymupdf.table.Table object at 0x000002A8EB257B10
(39.20347250075567, 151.95047505696616, 152.50390625, 165.32421875)

None

Error Log:

---

ValueError Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_30544\925744604.py in ?()
3 for i,tab in enumerate(tabs): # iterate over all tables
4 print(i,tab)
5 for cell in tab.header.cells:
6 print(cell)
----> 7 page.draw_rect(cell,color=fitz.pdfcolor["red"],width=0.3)
8 print("--")
9 page.draw_rect(tab.bbox,color=fitz.pdfcolor["green"])
10 print(f"Table {i} column names: {tab.header.names}, external: {tab.header.external}")

~.conda\envs\oolamaexp\Lib\site-packages\pymupdf\utils.py in ?(page, rect, color, fill, dashes, width, lineCap, lineJoin, morph, overlay, stroke_opacity, fill_opacity, oc, radius)
2159 '''
2160 Draw a rectangle. See Shape class method for details.
2161 '''
2162 img = page.new_shape()
-> 2163 Q = img.draw_rect(pymupdf.Rect(rect), radius=radius)
2164 img.finish(
2165 color=color,
2166 fill=fill,

~.conda\envs\oolamaexp\Lib\site-packages\pymupdf_init_.py in ?(self, p0, p1, x0, y0, x1, y1, *args)
11004
11005 Explicit keyword args p0, p1, x0, y0, x1, y1 override earlier settings
11006 if not None.
11007 """

11008 x0, y0, x1, y1 = util_make_rect( *args, p0=p0, p1=p1, x0=x0, y0=y0, x1=x1, y1=y1)
11009 self.x0 = float( x0)
11010 self.y0 = float( y0)
11011 self.x1 = float( x1)

~.conda\envs\oolamaexp\Lib\site-packages\pymupdf_init_.py in ?(p0, p1, x0, y0, x1, y1, *args)
20234 return args[0], args[1], x1, y1
20235 elif len(args) == 4:
20236 return args[0], args[1], args[2], args[3]
20237 raise Exception( f'Unrecognised args: {args}')

20238 ret_x0, ret_y0, ret_x1, ret_y1 = handle_args()
20239 if p0 is not None: ret_x0, ret_y0 = get_xy(p0)
20240 if p1 is not None: ret_x1, ret_y1 = get_xy(p1)
20241 if x0 is not None: ret_x0 = x0

ValueError: not enough values to unpack (expected 4, got 1)

[thangarajdeivasikamani]

After comment the below line. It's extract all the tables from respective page.

tabs = page.find_tables() # detect the tables
#print(tabs)
for i,tab in enumerate(tabs): # iterate over all tables
#print(i,tab)
for cell in tab.header.cells:
print(cell)
#page.draw_rect(cell,color=fitz.pdfcolor["red"],width=0.3)
page.draw_rect(tab.bbox,color=fitz.pdfcolor["green"])
print(tab.bbox)
print(f"Table {i} column names: {tab.header.names}, external: {tab.header.external}")
print("First table")
show_image(page, f"Table & Header BBoxes")

Is there any option to save the image as file instead of show_image in the last line.