Skip to main content

PDF catalogue import

Import products from PDF catalogues even the most complex layouts

Drop a supplier catalogue, price list or data sheet. WISEPIM reads its product tables, including merged cells, side-by-side columns, multilingual headers and scanned pages, and shows you every row before anything changes.

Start freeBook a demo
  • Up to 400 pages per PDF
  • Scanned pages read with OCR
  • Product photos extracted

Imported rows

Article numberdmm DBkg
7204B2047140,11
7204BETNG2047140,111
B7204AATB2047140,111
7304B2052150,14
7205B2552150,12
B7205AATB2552150,124
7205BETNG2552150,135

45 of 45 products, every value exact

Supplier PDF

See all 45 products and two more pages

Key features

  • Built for complex layouts

    Merged cells, wrapped column pairs, stacked headers and group headings become clean product rows.

  • Scanned pages too

    Pages without a text layer are read with built-in text recognition, and doubtful values are flagged.

  • Context kept

    Each row keeps its page, the section text above its table and the product photos beside it.

Real catalogue pages

Three hard pages, zero wrong values

These are real supplier PDFs, not samples we made to look good. Pick one to see the page next to the rows WISEPIM imported from it.

What makes this page hard

Pick a challenge to see where it sits on the page and what WISEPIM imported.

Polish catalogue page of single-row angular contact ball bearings: bore sizes 20 to 85 mm in merged yellow cells, dimensions, load ratings, speed limits, designations and mass
The PDF page
Polish ball bearing catalogue, 1 page, 45 products.

What WISEPIM imported

Text printed above the table, kept on every rowAABBETNGTBWymiaryNOŚNOŚCPrędkość granicznaOznaczenieMasa
Rows WISEPIM imported from this page, 8 of 45 shown
Article numberdmm DBrs min.kN CComin -1Column 8kg
7204B204714113,37,6412 00016 0000,11
7204BETNG204714113,37,6412 00016 0000,111
B7204AATB204714114,48,5812 00016 0000,111
7304B2052151,117,39,611 00015 0000,14
7205B255215114,89,310 00014 0000,12
B7205AATB255215113,68,110 00014 0000,124
7205BETNG255215115,89,8110 00014 0000,135
7305BETNG2562171,124,214,79 40011 0000,242

45 of 45 products, all 450 values exact

One imported record7204B

Row 1 of 45

Printed on the page

Article number
7204B
d
20
mm D
47
B
14
rs min.
1
kN C
13,3
Co
7,64
min -1
12 000
Column 8
16 000
kg
0,11
Section
AA | B | BETNG | TB | Wymiary | NOŚNOŚC | Prędkość graniczna | Oznaczenie | Masa

Added to every record

Page
3
Title
Łożyska skośne jednorzędowe ZVL SLOV AKIA, a. s.
Group
No group heading on this page
Notes
No notes on this page
Images
Cropped from the page on import, not shown in this demo
Needs review
Nothing flagged on this page
  • Checked value by value against hand-made truth tables
  • A test fails the build if one value changes
  • Headers kept as printed, matched to your fields on import

Layouts it reads

Made for the pages that break importers

Supplier catalogues are designed to be printed, not imported. These are the layouts WISEPIM turns into product rows without any setup.

  • Merged cells

    One value printed for several rows, like a thread size beside four article numbers, is copied to every row it covers. WISEPIM reads the drawn cell borders and shading, not just positions.

  • Side-by-side tables

    Price lists set in narrow column pairs, and two tables next to each other on one page, become one list of products.

  • Multilingual headers

    A description block printed in several languages becomes one section column per language, so every row carries all of them.

  • Fonts that decode wrongly

    Some PDFs use fonts that PDF readers decode wrongly, for example 7204Bb TNd. WISEPIM detects those words and reads them again.

  • Stacked headers and group headings

    Headers over two or three lines are joined, and a group heading between rows becomes a column on the rows below it.

  • Scanned pages

    Pages without a text layer are read with built-in text recognition. Values it is unsure of are flagged for review.

  • Product photos

    Photos printed beside a table are cropped from the page as you see it and attached to its products, up to four per table.

  • Product cards and line sheets

    Grids where each product is a small block of text under a photo are read as products too, with the shared details of the row.

  • Indexes and repeated articles

    A part number index is skipped. An article that appears in two tables, such as dimensions and prices, becomes one product with both.

Why not copy and paste

What a plain text copy loses

Copy the text of a catalogue page and the table falls apart. Values printed once for a group of rows end up on a line of their own, with nothing tying them to the products.

Plain text copy

Text layer of the page, extracted with pdftotext

             Wymiary                                             NOŚNOŚC                  Prędkość graniczna               Oznaczenie            Masa               mm                                                   kN                           min-1                      łożyska               kg                                                                  C                Co                                         rsd       D               B                                        min.         47              14               1               13,3              7,64           12 000           16 000             7204B               0,11        47              14               1               13,3              7,64           12 000           16 000           7204BETNG            0,11120Bore size, printed once        47              14               1               14,4              8,58           12 000           16 000           B7204AATB            0,111        52              15               1,1             17,3              9,6            11 000           15 000             7304B               0,14

WISEPIM rows

Article numberFilled from the merged celldmm DBkg
7204B2047140,11
7204BETNG2047140,111
B7204AATB2047140,111
7304B2052150,14

Copied text keeps the bore size on one line per group. The importer reads the drawn cell and puts 20 on all four rows.

No silent guesses

It flags doubts instead of guessing

A wrong value in a catalogue import is worse than a missing one. The importer only fills what the page prints and tells you where it was unsure.

  • Digits are never changed

    On scans, a misread separator or letter is restored to match the rest of its column. Digits are never added, dropped or changed.

  • Doubtful values are named

    Values the text recognition was unsure of are listed per product in a Needs review column, so you check those and nothing else.

  • You see it before it lands

    Every import opens on a review of the rows and matched columns. Nothing changes in your catalogue until you confirm.

How it works

From PDF to products in three steps

No templates, no zones to draw, no export to Excel first.

  1. 01

    Drop the PDF

    Open the importer and drop a catalogue, price list or data sheet of up to 64 MB and 400 pages.

    Product tables are found without any setup.

  2. 02

    Check the rows

    See each product as a row with its page, section and photos. The importer reports what it read, skipped and flagged.

    You know exactly what will be imported.

  3. 03

    Match columns and import

    Match the printed columns to your fields, then import. Values to check stay visible in the review.

    Products in your catalogue, ready to enrich.

Before you import

Check your PDF first

Five quick questions tell you whether a file suits the PDF importer. Nothing is uploaded, and you can still review every row before anything changes.

  1. Each product row has its own article, part or order number.

  2. The file is under 64 MB and 400 pages.

  3. No more than 150 pages are scans.

  4. Products sit in tables or product cards.

  5. The file lists several products, not one product drawing.

0 of 5 answered

FAQs

PDF import questions

What it reads, what it needs and where its limits are.

Still have questions?

Can't find the answer you're looking for? Please get in touch with our team.

Contact Support

Not every one. Supplier layouts vary. These three pages are checked value by value on every build. For your own file you see every row and every flagged value before anything is imported, and nothing changes until you confirm.

PDFs with product tables where each row has an article, part or order number: supplier catalogues, price lists, technical data sheets and line sheets. Text-layer PDFs and scanned pages both work.

Pages without a text layer are read with built-in text recognition. Misread separators are corrected to match their column, and values the recognition was unsure of are listed per product in a Needs review column.

Up to 64 MB and 400 pages per PDF, with up to 150 scanned pages read per file. Split larger catalogues into several files.

Rows without their own article or part number, pages of captioned photos and single-product drawings. A file without any product table is refused with the reason, so you never import empty rows.

Yes. Photos printed beside a table are cropped from the page and attached to its products, up to four per table.

No. You don't need to draw table areas, pick pages or convert to Excel. If a column has no printed name, it arrives as Column N and you name it when you match columns to your fields.

Turn your supplier PDFs into products

Create a workspace, open the file importer, drop your hardest PDF and review the rows. Nothing changes until you confirm.

Start free
Book a demo
4.7

Importing products

How file imports work in WISEPIM, from dropping a file to matching columns and reviewing the result.

(opens in a new tab)