From 8ac81c620527c8066503cd226f7a028c73ccc9ae Mon Sep 17 00:00:00 2001
From: Darek <darek.bidek@gmail.com>
Date: Wed, 16 Oct 2024 15:40:26 -0400
Subject: [PATCH] Setting inferSchemaLength = 0 when null (#285)

Setting inferSchemaLength = 0 when null to close #279
---
 __tests__/io.test.ts       | 10 ++++++++--
 polars/io.ts               | 11 +++++++++++
 polars/lazy/expr/string.ts |  2 +-
 polars/series/string.ts    |  2 +-
 4 files changed, 21 insertions(+), 4 deletions(-)

diff --git a/__tests__/io.test.ts b/__tests__/io.test.ts
index 7985dfce..14de9db3 100644
--- a/__tests__/io.test.ts
+++ b/__tests__/io.test.ts
@@ -24,7 +24,7 @@ describe("read:csv", () => {
     expect(df.shape).toEqual({ height: 27, width: 4 });
   });
   it("can read from a csv file with inferSchemaLength = 0 option", () => {
-    const df = pl.readCSV(csvpath, { inferSchemaLength: 0 });
+    let df = pl.readCSV(csvpath, { inferSchemaLength: 0 });
     const expected = `shape: (1, 4)
 ┌────────────┬──────────┬────────┬──────────┐
 │ category   ┆ calories ┆ fats_g ┆ sugars_g │
@@ -34,6 +34,8 @@ describe("read:csv", () => {
 │ vegetables ┆ 45       ┆ 0.5    ┆ 2        │
 └────────────┴──────────┴────────┴──────────┘`;
     expect(df.head(1).toString()).toEqual(expected);
+    df = pl.readCSV(csvpath, { inferSchemaLength: null });
+    expect(df.head(1).toString()).toEqual(expected);
   });
   it("can read from a csv file with options", () => {
     const df = pl.readCSV(csvpath, { hasHeader: false, skipRows: 1, nRows: 4 });
@@ -154,7 +156,11 @@ describe("read:json", () => {
     expect(df.shape).toEqual({ height: 27, width: 4 });
   });
   it("can specify read options", () => {
-    const df = pl.readJSON(jsonpath, { batchSize: 10, inferSchemaLength: 100 });
+    let df = pl.readJSON(jsonpath, { batchSize: 10, inferSchemaLength: 100 });
+    expect(df.shape).toEqual({ height: 27, width: 4 });
+    df = pl.readJSON(jsonpath, { batchSize: 10, inferSchemaLength: null });
+    expect(df.shape).toEqual({ height: 27, width: 4 });
+    df = pl.readJSON(jsonpath, { batchSize: 10, inferSchemaLength: 0 });
     expect(df.shape).toEqual({ height: 27, width: 4 });
   });
   it("can read from a json buffer", () => {
diff --git a/polars/io.ts b/polars/io.ts
index 7c02b960..37ea88ef 100644
--- a/polars/io.ts
+++ b/polars/io.ts
@@ -186,6 +186,9 @@ export function readCSV(pathOrBody, options?) {
   options = { ...readCsvDefaultOptions, ...options };
   const extensions = [".tsv", ".csv"];
 
+  // Handle If set to `null` case
+  options.inferSchemaLength = options.inferSchemaLength ?? 0;
+
   if (Buffer.isBuffer(pathOrBody)) {
     return _DataFrame(pli.readCsv(pathOrBody, options));
   }
@@ -275,6 +278,8 @@ export function scanCSV(
 export function scanCSV(path, options?) {
   options = { ...scanCsvDefaultOptions, ...options };
 
+  // Handle If set to `null` case
+  options.inferSchemaLength = options.inferSchemaLength ?? 0;
   return _LazyDataFrame(pli.scanCsv(path, options));
 }
 /**
@@ -320,6 +325,10 @@ export function readJSON(
   options = { ...readJsonDefaultOptions, ...options };
   const method = options.format === "lines" ? pli.readJsonLines : pli.readJson;
   const extensions = [".ndjson", ".json", ".jsonl"];
+
+  // Handle If set to `null` case
+  options.inferSchemaLength = options.inferSchemaLength ?? 0;
+
   if (Buffer.isBuffer(pathOrBody)) {
     return _DataFrame(pli.readJson(pathOrBody, options));
   }
@@ -382,6 +391,8 @@ export function scanJson(
 export function scanJson(path: string, options?: Partial<ScanJsonOptions>) {
   options = { ...readJsonDefaultOptions, ...options };
 
+  // Handle If set to `null` case
+  options.inferSchemaLength = options.inferSchemaLength ?? 0;
   return _LazyDataFrame(pli.scanJson(path, options));
 }
 
diff --git a/polars/lazy/expr/string.ts b/polars/lazy/expr/string.ts
index f4718f55..239fa4d8 100644
--- a/polars/lazy/expr/string.ts
+++ b/polars/lazy/expr/string.ts
@@ -164,7 +164,7 @@ export interface StringNamespace extends StringFunctions<Expr> {
    * @see https://goessner.net/articles/JsonPath/
    * @param jsonPath - A valid JSON path query string
    * @param dtype - The dtype to cast the extracted value to. If None, the dtype will be inferred from the JSON value.
-   * @param inferSchemaLength - How many rows to parse to determine the schema. If ``None`` all rows are used.
+   * @param inferSchemaLength - How many rows to parse to determine the schema. If `null` all rows are used.
    * @returns Utf8 array. Contain null if original value is null or the `jsonPath` return nothing.
    * @example
    * ```
diff --git a/polars/series/string.ts b/polars/series/string.ts
index a2ad7baa..a03b4d07 100644
--- a/polars/series/string.ts
+++ b/polars/series/string.ts
@@ -132,7 +132,7 @@ export interface StringNamespace extends StringFunctions<Series> {
    * @see https://goessner.net/articles/JsonPath/
    * @param jsonPath - A valid JSON path query string
    * @param dtype - The dtype to cast the extracted value to. If None, the dtype will be inferred from the JSON value.
-   * @param inferSchemaLength - How many rows to parse to determine the schema. If ``None`` all rows are used.
+   * @param inferSchemaLength - How many rows to parse to determine the schema. If ``null`` all rows are used.
    * @returns Utf8 array. Contain null if original value is null or the `jsonPath` return nothing.
    * @example
    * ```