CSV exports

Load a folder of Screaming Frog CSV exports using the CSV backend — no Java, no database, no setup required.

The CSV backend reads a directory of exported .csv files that you have already produced from the Screaming Frog UI (or CLI). It is the simplest way to get started if you already have exports on disk.

Loading from a folder

from screamingfrog import Crawl

crawl = Crawl.load("./exports")

Crawl.load detects a CSV directory automatically when the path is a folder containing .csv files. You can also call the constructor directly:

crawl = Crawl.from_exports("./exports")

Listing available tabs

Use crawl.tabs to see which CSV files are present in the folder:

print(crawl.tabs)
# ['internal_all', 'response_codes_all', 'page_titles', ...]

Tab names match the CSV filename without the .csv extension, normalized to lowercase with underscores.

Accessing tabs

Access any tab by its normalized name:

for row in crawl.tab("response_codes_all"):
    print(row["Address"], row["Status Code"])

You can also pass the full filename — the extension is optional:

rows = crawl.tab("page_titles.csv").collect()

Filtering rows

# Filter by column value
for row in crawl.tab("internal_all").filter(status_code="404"):
    print(row["Address"])

# Apply a GUI filter (matches the Screaming Frog UI filter name)
for row in crawl.tab("page_titles").filter(gui="Missing"):
    print(row["Address"], row["Title 1"])

Page and link views

The first-class pages() and links() views work with the CSV backend too:

pages_404 = crawl.pages().filter(status_code=404).collect()
nofollow_links = crawl.links("in").filter(rel="nofollow").collect()

crawl.links("in") requires an all_inlinks.csv file in the export folder. If that file is missing, the view will be empty.

Limitations

• No raw SQL. crawl.raw(), crawl.sql(), and crawl.query() are not supported on the CSV backend.
• Filtered CSVs match GUI output. When you export with a GUI filter active in Screaming Frog, the CSV contains only the filtered rows. This means crawl.tab("page_titles") returns whatever rows were in the file — it does not re-apply the filter.
• No Derby-mapped fields. Computed fields like Indexability and Indexability Status that Derby derives from raw columns are not available unless those columns appear in your exported CSVs.
• Coverage depends on what you exported. Only the tabs you exported from the UI are available. Use export_profile="kitchen_sink" with the CLI to produce a broad export.

When to use the CSV backend

Use CSV when

• You already have exports on disk and do not want to open Derby.
• You need exact GUI filter output (e.g., Page Titles > Missing).
• You are on a machine without Java.
• Your workflow is simple: read rows, filter, export.

Use Derby / DuckDB when

• You need raw SQL access to Derby tables.
• You want to query tabs that you did not export.
• You need computed fields like Indexability or Indexability Status.
• You are doing large-scale analysis and want DuckDB’s columnar performance.

Tab metadata helpers

# List GUI filter names for a tab
print(crawl.tab_filters("Page Titles"))

# Inspect column headers
print(crawl.tab_columns("page_titles"))

# Get filters and columns together
print(crawl.describe_tab("page_titles"))