Profile every column of a CSV — type, empty rate, unique count, min/max — then de-duplicate and sort it.
逐列统计 CSV 的数据类型、空值率、去重数量与最值，并一键去重、排序、导出干净数据。
A CSV profiler that tells you what is actually inside a file before you process it. It parses the CSV (with quoted fields, embedded commas and escaped quotes handled correctly), then reports for every column: the inferred type — text, number or date — how many values are empty, how many distinct values exist, and the minimum and maximum for numeric columns. From there you can de-duplicate rows, sort by a column, and export the cleaned file.
一个 CSV 剖析工具，在动手处理之前先告诉你文件里到底有什么。它按 RFC 4180 正确解析 CSV（引号字段、内嵌逗号、转义引号都支持），然后逐列报告：推断出的类型——文本、数字或日期——空值数量、去重后的取值个数，数值列还会给出最值。接着可以一键去重、按列排序，并导出清洗后的文件。
A column is numeric when every non-empty value parses as a number, and date-like when its first value starts with a YYYY-MM-DD pattern. Everything else is text. Empty cells never force a type.
当一列所有非空值都能解析成数字时判定为数字型；首个值形如 YYYY-MM-DD 时判定为日期型；其余为文本。空单元格不会影响类型判定。
Yes. The parser is quote-aware: commas, line breaks and doubled quotes inside a quoted field are handled the way RFC 4180 specifies, and the exported file quotes those fields again.
能。解析器是引号感知的：引号内的逗号、换行和双写引号都按 RFC 4180 规则处理，导出时也会重新给这些字段加引号。
Yes. The first occurrence of each unique row is kept and the duplicates that follow it are dropped.
会。每组重复行保留第一次出现的行，其后的重复行被丢弃。
No. Parsing and profiling happen in the browser.
不会。解析与统计都在浏览器内完成。
Open an .xlsx or .csv file, preview the sheet, then export it as…
打开 .xlsx 或 .csv 文件预览表格内容，再导出为 JSON、CSV 或…
Load a .sqlite database and run SQL queries against it entirely …
载入 .sqlite 数据库，用 WASM 版 SQLite 引擎在浏览器里执行…
Query and reshape JSON with JSONata expressions, list every path…
用 JSONata 表达式查询与重塑 JSON 数据，列出全部数据路径，或把嵌套…