Tablora · Library
Inspect records with CRLF and LF endings
Text pasted from several exports can mix CRLF, LF and CR line endings. The inspector treats each as a record boundary unless it sits inside quotes.
Where mixed line endings come from
Windows programs usually end each line with a carriage return and a line feed (CRLF), Unix and macOS tools with a line feed alone (LF), and some older systems with a lone carriage return (CR). When you join the output of different programs, or paste part of a file into another, a single block of text can contain all three. Some tools then merge two records into one or split one record into two.
The inspector accepts all three as the end of a record when they are outside a quoted field. A line break inside quotes is part of that cell's text and does not start a new record.
Counting the records
- Paste the complete text into Table text in one go rather than in pieces.
- Choose the Delimiter or leave Detect delimiter selected, and press Calculate.
- Compare Rows with the number of records you expect from the source pages.
- Check Rows with a different width; a value of 0 means every record has the same number of fields as the header.
- If the data has quoted notes, inspect those cells separately, since their internal line breaks are kept.
The Mira and Noah example
The page example has a header id,name and two records, 1,Mira and 2,Noah, with comma as the Delimiter and different line endings between the lines. The inspector reports 3 rows and 2 columns. Row 3, column 2 reads Noah, which shows that the last record was found even though the line endings were not consistent.
A single line break at the very end of the text does not add an empty record.
What the download does to line endings
Download (CSV) writes every record boundary as CRLF, whatever the input used, so the downloaded file is consistent. Line breaks inside a quoted cell are written as they were, inside quotes. The Tablora app's CSV export also ends every row with CRLF.
If the receiving program still shows the wrong number of records, check its import settings for quoted fields before changing the data.
When a count still looks wrong
- More rows than expected can mean a line break inside a value that was not quoted in the source.
- Fewer rows than expected can mean an opening quote that swallows following lines; look for an unclosed quoted field.
- Do not replace line endings by hand before inspecting the text; the parser already handles all three.
- Text from a system that ends lines with a lone carriage return can look like one long line in some editors. The inspector still splits it into records, so trust Rows over what the editor shows.
Original synthetic example
- Input
- id,name 1,Mira 2,Noah
- Delimiter
- ,
- Expected records
- 3
- Columns in the first record
- 2
- Cell to inspect
- Row 3, column 2: Noah