OmniDocHow to extract only the required part of the document
When you only need a certain section, sentence or several fields in the document, how can you quickly find and extract them individually? This article discusses four methods of copying, searching and positioning, splitting by page, and cropping after exporting based on target classification. It also provides judgments before starting and a checklist after completion.
Key points of this article
- First determine whether you want to extract a whole section, a whole paragraph, a sentence or a certain field. Different methods are used at different granularities.
- The text version of the PDF can be directly selected and copied; the scanned version of the PDF must be OCRed first, and then the text can be extracted.
- When extracting by page, first confirm the page number range in the file, and then use the page number and page number to check whether all files have been extracted.
- The extracted content should be compared with the original text, especially figures, tables and proper nouns.
Many people open a report, contract or paper with dozens of pages and only want the conclusion, a clause or a set of figures in one section, but they have to flip through the entire document. How to extract part of the content from a document is a very common problem in daily work, but the methods are quite different for different file types.
This article is written for readers who need to extract specified content from Word, PDF or a long text. After reading, you will be able to judge which method you should use, know what to confirm before taking action, and use the OmniDoc desktop client to complete operations such as splitting by page and sorting after exporting. Finally, there will be a checklist to avoid omissions or errors.
The following methods are not limited to a certain software. The core idea is: first clarify what you want to retrieve, then choose a method that retains the location information of the original text, and finally check item by item.
How to extract part of the content from the document? One sentence conclusion
In most cases, choosing the right method is more important than finding the right tool. The table below provides methods based on your situation.
| your situation | Recommended practices | reason |
|---|---|---|
| As long as one or two paragraphs or one sentence is required, the document is a text version | Select it in the reading interface, copy it, and paste it into a new document. | Fast operation, no need to process the entire file |
| The file is very long and the target content is scattered in many places | First use search to locate keywords, mark them one by one, and then copy them | Avoid scrolling repeatedly throughout the document |
| A full page or several consecutive pages are required, and the layout must be retained | Split by page range in PDF editing, save as new file | Layout and page content remain intact |
| Editable content is required, the original text is PDF | Export to Word or Excel first, then delete the redundant parts | You can modify it directly after exporting, and the table is easier to organize. |
| The original text is a scanned copy and the text cannot be selected. | Do OCR recognition first, then select or export for extraction | Only the recognized text layer can be selected and searched |
The key to judgment is: whether you want "position" (certain pages, a certain section) or "content" (certain fields, certain sentences). The position is suitable for splitting pages; the content is suitable for sorting after copying or exporting.
Before you start: Confirm these points first
Taking a minute to confirm the following items before starting can save a lot of rework.
- Whether the document is a text version or a scanned version. Try dragging the mouse to select a text in the PDF: if it can be selected and highlighted, it means it is a text version; if it cannot be selected or it can only select the entire picture, it is probably a scanned copy. The scanned document needs to be OCRed first, otherwise the text cannot be extracted.
- The granularity to extract. Is it a whole section, a paragraph, a sentence, or some cells in the table? The finer the granularity, the more manual checking is required after exporting, because formatting and line breaks are prone to errors when copying.
- Whether the original format needs to be retained. If you are just excerpting the text, just copy and paste it; if you want to give it to others to see the original page, you should split it by page and retain the layout of the original document.
- Whether the original file should be retained. No matter which method is used, it is recommended to make a copy of the original file before operating. Splitting, exporting, and deleting content are all done on copies, and the originals are retained. If you make a mistake, you can go back and extract it again.
- Use after extraction. If the content is to be used for citations, quotations or legal documents, it must be checked verbatim; if it is only for internal reference, the verification can be relaxed appropriately, but the key figures must still be read again.
Steps to extract part of PDF content by page in OmniDoc
The following steps apply when you need to remove consecutive pages. If all you want is a few paragraphs of text, you can skip splitting and select and copy directly in the PDF.
- Download and install the OmniDoc desktop client at download.omnidoc.top, which supports Windows, macOS and Linux.
- Open the PDF to be processed. If it is a scanned document, first use the OCR function to recognize the text so that the document has a selectable text layer.
- Find the split function in the page-related operations and specify the parts to keep according to the page number range, such as "Page 12 to Page 18".
- Confirm that the split result is a new file and the original file remains unchanged to avoid overwriting the original.
- If you need to modify the text, directly modify the paragraph text in the PDF editing function, or use the export function to convert the content to Word or Excel.
- After exporting, delete redundant paragraphs or tables in Word, save it as the final version, and check it item by item in the checklist in the next section.
If you are already working in a Word document, you can also directly extract the content in Word by selecting, copying, and searching. The process is similar to the above. OmniDoc's UniDoc web page editor (app.unidoc.top) also supports Word-compatible editing, where exported content can be organized.
Check each item after completion
Once the extraction is complete, don't rush to deliver or reference. The checklist below can help you identify common problems.
| Check items | How to see | What to do if there is a problem |
|---|---|---|
| Page number or starting and ending position | Check the original text table of contents or page numbers to confirm that the first and last pages are correct. | Reset the page number range and split it again |
| Whether content is missed | Search the original text for a few words at the beginning and end of the sentence to see if they are all found. | Add missing paragraphs or reselect from the original file |
| Numbers and units | Compare the numbers, decimal points, percent signs and units in the original text one by one | Correct based on the original text, do not modify based on impressions |
| Table structure | Check whether the number of rows and columns, merged cells and table headers are misaligned | Reformat in Excel format, or rebuild the table manually instead |
| Scanned document recognition results | Spot-check whether names, proper nouns and long numbers are recognized correctly | Manually correct errors and compare important content verbatim with the original. |
| Machine translation or conversion traces | Check for garbled characters, extra spaces and line breaks | Remove redundant characters and unify paragraph format |
Both OCR and format conversion have recognition or restoration errors. The extracted content, especially numbers and proper nouns, should be checked again with the original file.
❓ Frequently Asked Questions (FAQ)
1. Can text be selected directly from the scanned PDF?
Most scanned PDFs are just pictures, and Chinese characters cannot be selected directly. First use OCR recognition to generate a text layer, and then you can select, search and copy. After the identification is completed, key contents still need to be spot-checked.
2. Will the format change after extraction?
When text is copied and pasted, the formatting is often lost, and paragraphs and tables are easily confused. If you need to preserve the layout, you should split it into new PDFs by page; if you need to edit, you can export it to Word first and then organize it.
3. How to quickly find the target content in a long document?
Use the search function in the reading interface, enter keywords or a few words in the sentence, and locate them one by one. After you find it, write down the page number first, and then select or split it together, which is faster than searching page by page.
4. I only want to extract certain fields in the document. Is there any way?
Fields are usually scattered among tables or fixed-format paragraphs. You can first export the file to Excel or Word, then use the filter or search function of the table to locate the fields, and finally manually copy it to the target location. After extraction, check whether the field names and corresponding values correspond one-to-one.
📝 Summary
How to extract part of the content from the document, the key is to first distinguish whether it is the location or the content, and then choose the corresponding method. The content of the text version can be selected and copied. The entire page is suitable for splitting by page. When editing is needed, export it first and then organize it. After the extraction is completed, check it one by one according to page numbers, figures and tables before you can use it with confidence.
