OmniDocHow to extract PDF tables while keeping the original format unchanged
When extracting a table from PDF, first determine whether it is a text version or a scanned version, and then select the export method. After exporting, check the rows, columns, numbers, merged cells, and borders one by one. This article explains the settings and checks to be done in advance.
Key points of this article
- First use the mouse to select a cell text in the table, determine whether the PDF is a text version or a scanned version, and then decide on the extraction method.
- The text version is first exported to Excel or Word, which can better preserve the row and column structure and correspondence than direct copy and paste.
- The scanned version must be OCR recognized first, otherwise only images will be exported, and the table cannot remain in editable format.
- After exporting, check the number of rows and columns, numbers, merged cells, borders and cross-page headers one by one, and compare them with the original table.
- Merged cells and diagonal headers are easily misplaced. Please check them manually when encountering them. Do not use important numbers directly.
After copying a table from PDF and pasting it into Excel or Word, a common situation is that the column widths are misaligned, the borders disappear, and the numbers move into other cells. In the end, you can only manually adjust each cell. Financial statements, experimental data, and contract quotations are all prone to this problem. This article is for those who need to move forms from PDF to somewhere else to continue using them.
This article talks about "what to do beforehand": before the table leaves the PDF, choose the right extraction method and export method, and try to preserve the alignment, borders and fonts. How to repair a messy table is another type of question. I will only mention it here at the end.
After reading, you can judge which extraction method is suitable for the PDF in your hand, know what things to confirm before starting, and use several checks to quickly find problems after exporting.
How to extract PDF tables so that the format remains unchanged? One sentence conclusion
First determine whether the PDF is a text version or a scanned version. The text version will give priority to the export method that can preserve the table structure; the scanned version must be OCR recognized first and then exported. It's difficult to preserve borders, column widths, and alignment at the same time by just copying and pasting.
| your situation | Recommended practices | reason |
|---|---|---|
| Text version PDF, simple table structure, no merged cells | Export to Excel or Word using PDF editor | The row and column structure and numbers can be mapped to the correct grid |
| Text version PDF, with a large number of merged cells or slash headers | After exporting, check table by table and manually correct it in the target software if necessary. | Automatic recognition of merged cells is prone to errors and requires manual comparison. |
| Scanned PDF, the content cannot be selected when selecting text | Use OCR to identify first, then export | When there is no text layer, you can only get pictures when exporting directly. |
| Only the data in the table is needed, and the format can be rearranged | After exporting to Excel, only keep the numerical columns, or copy the plain text and then divide it into columns. | When formatting is not important, adjustment time can be significantly reduced |
| The content of the form needs to be translated into Chinese and English. | Upload PDF on omnidoc.top, select bilingual comparison or translation only and download | Tables, layouts and formulas are preserved during translation |
Before you start: Confirm these points first
- Confirm whether the PDF is a text version or a scanned version. Use the mouse to drag and select a cell text in the table. If it can be selected and highlighted, it means there is a text layer; if it cannot be selected, or the entire picture is selected, it is a scan or a picture table. The processing paths of the two are different, and if you make a wrong judgment, your efforts will be in vain.
- Count the complexity of the table. Check to see if there are merged cells, multi-layer headers, and long tables that span pages. The more complex the structure, the easier it is for automatic extraction to be misaligned, and more time needs to be reserved for verification.
- Make it clear what to use after exporting. If you still need to calculate or filter, exporting to Excel is more suitable; if you also need to typeset, print, or submit into a document, exporting to Word is more suitable. The target software determines which format of the table structure is easier to maintain.
- Check the quality of the source PDF itself. For example, whether the page has been rotated, whether the table has been cut off by page breaks, and whether the scanned document is skewed or blurred. If there is a problem with the source file, there will also be a problem with the exported result. Dealing with the source first can save a lot of subsequent repairs.
- Keep a copy of the original document. There may be errors in both export and recognition. You need to check the original table at any time when checking, so don't just save the exported version.
Steps to export tables in OmniDoc desktop client
The following steps use the OmniDoc desktop client's PDF editor. The client supports Windows, macOS and Linux and can be downloaded from download.omnidoc.top.
- Open download.omnidoc.top, download and install the desktop client suitable for your system.
- Open the PDF file from which the table needs to be extracted in the client.
- Enter the PDF editor, turn to the page where the table is located, and confirm that the entire table is complete and is not cut across pages.
- If it is a scanned version, first use the OCR function to recognize the text. After the recognition is completed, select the cell text again to confirm that the text layer has been generated.
- Select Export to Excel or Word and decide where to save it.
- Open the exported file, first check the column width, table header and number position, and then compare it line by line with the original PDF.
- If the table is long or spans two pages, turn to each spread and check whether the table headers are repeated and whether the data is continuous.
- After confirming that it is correct, save it as the file for final use, and both the original PDF and the exported file will be retained.
Check each item after completion
Do not use it directly after exporting. Please go through the following check items. Most problems can be found at this step.
| Check items | How to see | What if there's a problem? |
|---|---|---|
| Number of rows and columns | Compare the number of rows and columns of the original table | If the rows or columns are missing, go back to the source PDF to check if there are hidden or merged squares, and fill them up manually. |
| Numbers and decimal places | Spot check key values such as amount, percentage, etc. | Digital dislocation or loss of decimal point, re-export, or direct correction in the target software |
| Merge cells | See if the merged places in the original table are still merged. | If it is taken apart, remerge it with the table software and make sure that the content is placed in the correct grid. |
| Font and alignment | See if the title is bold and whether the numbers are aligned uniformly. | Uniformly format the whole column without having to adjust it one by one |
| Frame | See if the outer frame and the internal separation line are complete. | Draw the frame when the line is missing, or rearrange it according to the original table structure. |
| Spread header | Turn to the next page to see if the header is repeated. | Manually fill in the header at the top of each page without repetition |
Whether it is OCR recognition or automatic export, the machine result may be typographical error, line error or digital misjudgment. The contents involving amount, reimbursement, audit or external submission must be checked with the original item by item before use.
# Core comparison
There are three common extraction methods, each of which has different format retention. You can refer to the following table when choosing.
| Method | Border and alignment | Numbers and words | Common risk |
|---|---|---|---|
| Copy and paste directly | The border is usually lost and the column width needs to be readjusted. | Text version is available, but numbers are easily squeezed into the same column | The format is easy to be messed up, and the complex table almost needs to be retyped. |
| Export to Excel | The row and column structure is better retained, and the border depends on the original table. | The numerical value can be divided into the corresponding grid. | Merging cell and slash headers is error prone |
| Export to Word | The layout and table appearance are close to those of the original document. | Suitable for continued typesetting, printing or submission | After a long table spreads, you may need to adjust the header and paging |
| OCR before export | Depends on the quality of recognition and the clarity of the original table. | It is a necessary pre-step for scanning parts. | Wrong words and lines may occur, which need to be checked item by item. |
# frequently asked questions (FAQ)
1. Why is it all messed up after copying and pasting the PDF form directly?
PDF mainly records the position of each word on the page, not the column structure of the table. When copying, the software has to re-determine which words belong to the same grid, and if the judgment is wrong, it will be misplaced. Exporting to Excel or Word is usually more stable than pasting directly.
2. Can the scanned PDF form remain in its original format?
The scanned piece has no text layer and needs to be identified by OCR first. How much the format can be retained after recognition depends on the clarity and structural complexity of the original table, and the results still have to be checked item by item.
3. The form is very long and spans several pages. How to avoid format breakage?
Make sure there is a header on each page of the original table before exporting. After export, turn to each spread to check, fill the meter head when you find the break, and close the cut lines and go back.
4. Just want the data in the table, not the format, how to do it?
After exporting to Excel, you can keep only the numeric columns you need, or you can copy plain text and sort it out with the sorting function. This saves time when the format is not important.
5. Online conversion website extraction form, what should be paid attention to?
When selecting an online conversion site, first look at the file size limit, whether the export result is watermarked, whether the privacy policy indicates the retention period of the uploaded file, and whether the table structure can be retained. For sensitive financial or contract content, it is recommended to give priority to local processing.
# Summary
The key to PDF table extraction and retention format is to first judge the text version or the scanned version, and then choose the right export method. Confirm the complexity and use of the form before you start, and then check it one by one according to rows, numbers, merged cells, borders and spread headers, so that the risk of format confusion can be reduced.
