About the Handiwork CSV Viewer & Analyzer
The CSV Viewer & Analyzer opens delimited text files locally and turns them into a searchable, sortable table. It can detect common separators, decode several text encodings, identify uneven rows, infer basic column types, count empty and unique values, and calculate summaries for consistently numeric columns.
How to use the Handiwork CSV Viewer & Analyzer
- Choose or drop a CSV, TSV, or TXT file up to 15 MB.
- Confirm the detected delimiter, text encoding, and whether the first row contains column names.
- Search all analyzed rows, select column headings to sort, and use the page controls to inspect the table.
- Review the column analysis for inferred types, missing values, unique values, and numeric minimum, maximum, and average.
Quoted fields and delimiter detection
The parser recognizes double-quoted fields, escaped double quotes, and line breaks inside quoted fields. Automatic detection compares comma, tab, semicolon, and pipe separators outside quoted text, then chooses the candidate that appears most consistently across the first records. You can override the result at any time.
What the column analysis measures
Each analyzed column reports non-empty, empty, and distinct non-empty value counts. Values are classified as numbers, booleans, ISO-style dates, text, mixed, or empty. Minimum, maximum, and average appear only when every non-empty value in that column passes the viewer’s strict numeric check.
Handling different text encodings
CSV does not carry a universal encoding declaration inside the file. UTF-8 is the default, with optional Windows-1252, ISO-8859-1, and UTF-16 little-endian decoding for files exported by older systems. If characters look incorrect, choose another encoding and the browser will decode the original file bytes again.
Bounded analysis for browser stability
The tool accepts files up to 15 MB and analyzes at most 10,000 data rows and 100 columns. It still scans the file to report the total record and column counts, then identifies content beyond the analysis window. Table pages contain 100 rows so the browser does not create thousands of visible elements at once.
Assumptions and limitations
- Automatic delimiter detection considers comma, tab, semicolon, and pipe only; choose the separator manually if detection is wrong.
- Type inference is descriptive rather than schema validation. Identifiers, localized dates, currency strings, and numbers containing grouping separators may be classified as text.
- Only the first 10,000 data rows and 100 columns are searchable, sortable, and included in column statistics.
- The parser uses double quotes as the field qualifier and does not support configurable escape characters or multi-character delimiters.
- Changing the displayed encoding cannot repair bytes that were already corrupted before the file was created.
Sources and standards
These authoritative references were used to verify the method and guidance on this page.
Frequently asked questions
Is my CSV file uploaded?
No. The browser reads and analyzes the selected file locally. Handiwork does not send its contents to a server.
Can it open tab- or semicolon-separated files?
Yes. Automatic detection checks comma, tab, semicolon, and pipe separators, and the delimiter menu lets you override the detected choice.
Why are some rows marked inconsistent?
An inconsistent row has a different number of fields from the header row, or from the first data row when headers are disabled. Quoting or delimiter problems are common causes.
How are column types inferred?
The analyzer checks every non-empty value in the analyzed range. A column receives one type only when those values consistently match numbers, true/false booleans, ISO-style dates, or text; otherwise it is marked Mixed.
Does the analyzer change my file?
No. Searching, sorting, encoding changes, and analysis affect only the in-memory view. The source file is never modified.
What happens with a CSV containing more than 10,000 rows?
The tool reports the total record count but limits the searchable table and statistics to the first 10,000 data rows. A notice shows how many rows were excluded.