Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/exegia/corpora-py/llms.txt

Use this file to discover all available pages before exploring further.

search_csv() runs a Context-Fabric query template and writes the full result set — every matched node from every result tuple — to a CSV file on the filesystem where the server process is running. Each row represents one node from one result tuple and includes its result index, slot index within that tuple, numeric node ID, section reference, and one column per requested feature. This is the right tool when you want to post-process or visualise results outside the MCP session.
search_csv() writes directly to the local filesystem of the machine running the MCP server. It only works when the server was started with stdio transport (the default cf-mcp invocation without --sse or --http). It will not write files to a client machine.

​
Parameters

​
template
string
required
The query template string, using the same syntax as search(). See search_syntax_guide() for details.
​
output_path
string
required
Absolute path for the output .csv file on the server’s filesystem (e.g. "/home/user/results.csv"). The file is created or overwritten. The path is expanded for ~.
​
features
string[]
List of feature names to include as columns in the CSV. Defaults to ["otype", "lex", "pos", "gloss"] when omitted. Use list_features() to discover valid names for your corpus.
​
corpus
string
Corpus name. When omitted, the currently active corpus is used.

​
Returns

A confirmation string reporting how many results and total node rows were exported, and the resolved output path: 
Exported 4,892 results (4,892 rows) to /home/user/results.csv
Returns "No results found." when the template matches nothing. Returns "Search error: <message>" if the template is invalid.

​
CSV schema

ColumnTypeDescription
result_indexintegerZero-based index of the matched result tuple
slot_indexintegerZero-based position of this node within the tuple
nodeintegerNumeric node ID
sectionstringHuman-readable section reference (e.g. "Genesis 1 1")
(features)stringOne column per name in features, value from that feature for this node

​
Example

result = search_csv(
    template="word pos=verb",
    output_path="/tmp/verbs.csv",
    features=["lex", "pos", "gloss", "nu", "vt"],
)
print(result)
Expected output: 
Exported 34,611 results (34,611 rows) to /tmp/verbs.csv
The generated CSV will look like: 
result_index,slot_index,node,section,lex,pos,gloss,nu,vt
0,0,1234,Genesis 1 1,BRʾ[,verb,create,sg,perf
1,0,1289,Genesis 1 2,HJH[,verb,be,sg,wayq
...

Build docs developers (and LLMs) love

Get started for freeTalk to us