Segment-Wise Content Case Normalization

Hard

A publishing platform stores raw user-authored copy in a relational table and needs consistent title-style formatting without changing punctuation or spacing. Table: user_content content_id: unique row identifier content_text: original text payload Task: Transform each content_text value with the following rules: A segment is a maximal contiguous run of English letters [A-Za-z]. In every segment, uppercase the first letter and lowercase the remaining letters. Keep all non-letter characters exactly as they appear. Any non-letter character starts a new segment boundary for the next letter. Examples of boundary behavior: "modern-day" -> "Modern-Day" "api@GATEWAY" -> "Api@Gateway" "alpha/beta" -> "Alpha/Beta" Multiple spaces and repeated separators must be preserved exactly. Output requirements: Return columns in this exact order: content_id original_text converted_text Final ordering: content_id ascending Supported submission environments: SQL (SQLite 3.27.2) Pandas (Python)

Example 1

Input

user_content:
| content_id | content_text                    |
|------------|---------------------------------|
| 1          | hello world of SQL              |
| 2          | the QUICK-brown fox             |
| 3          | modern-day DATA science         |
| 4          | web-based FRONT-end development |

Output

[
  {"content_id":1,"original_text":"hello world of SQL","converted_text":"Hello World Of Sql"},
  {"content_id":2,"original_text":"the QUICK-brown fox","converted_text":"The Quick-Brown Fox"},
  {"content_id":3,"original_text":"modern-day DATA science","converted_text":"Modern-Day Data Science"},
  {"content_id":4,"original_text":"web-based FRONT-end development","converted_text":"Web-Based Front-End Development"}
]

Explanation

Each letter segment is independently normalized while spaces and punctuation stay unchanged.

Example 2

Input

user_content:
| content_id | content_text                          |
|------------|---------------------------------------|
| 10         |   mIXed   SPACING  and-HYPHEN   text  |
| 11         | alpha@@BETA//gamma^^delta             |

Output

[
  {"content_id":10,"original_text":"  mIXed   SPACING  and-HYPHEN   text  ","converted_text":"  Mixed   Spacing  And-Hyphen   Text  "},
  {"content_id":11,"original_text":"alpha@@BETA//gamma^^delta","converted_text":"Alpha@@Beta//Gamma^^Delta"}
]

Explanation

Repeated separators do not collapse. Each separator resets segment capitalization for the next letter.

Example 3

Input

user_content:
| content_id | content_text                    |
|------------|---------------------------------|
| 21         | pre,POST^merge\rollback       |
| 22         | one-two/THREE@four,five         |

Output

[
  {"content_id":21,"original_text":"pre,POST^merge\rollback","converted_text":"Pre,Post^Merge\Rollback"},
  {"content_id":22,"original_text":"one-two/THREE@four,five","converted_text":"One-Two/Three@Four,Five"}
]

Explanation

Commas, slashes, carets, and backslashes are preserved while still splitting letter segments.

Constraints

1 <= number of rows in user_content <= 200000
content_id is unique
1 <= content_id <= 1000000000
1 <= length(content_text) <= 400
content_text contains only English letters and these separators: space, hyphen, @, slash, caret, comma, backslash
All separators must remain exactly unchanged in output
Segment casing is applied independently between separators
SQL submissions run on SQLite 3.27.2
Pandas submissions must return columns [content_id, original_text, converted_text]

Preparing your editor. You can read the problem while it loads.

Segment-Wise Content Case Normalization

Hard

A publishing platform stores raw user-authored copy in a relational table and needs consistent title-style formatting without changing punctuation or spacing. Table: user_content content_id: unique row identifier content_text: original text payload Task: Transform each content_text value with the following rules: A segment is a maximal contiguous run of English letters [A-Za-z]. In every segment, uppercase the first letter and lowercase the remaining letters. Keep all non-letter characters exactly as they appear. Any non-letter character starts a new segment boundary for the next letter. Examples of boundary behavior: "modern-day" -> "Modern-Day" "api@GATEWAY" -> "Api@Gateway" "alpha/beta" -> "Alpha/Beta" Multiple spaces and repeated separators must be preserved exactly. Output requirements: Return columns in this exact order: content_id original_text converted_text Final ordering: content_id ascending Supported submission environments: SQL (SQLite 3.27.2) Pandas (Python)

Example 1

Input

user_content:
| content_id | content_text                    |
|------------|---------------------------------|
| 1          | hello world of SQL              |
| 2          | the QUICK-brown fox             |
| 3          | modern-day DATA science         |
| 4          | web-based FRONT-end development |

Output

[
  {"content_id":1,"original_text":"hello world of SQL","converted_text":"Hello World Of Sql"},
  {"content_id":2,"original_text":"the QUICK-brown fox","converted_text":"The Quick-Brown Fox"},
  {"content_id":3,"original_text":"modern-day DATA science","converted_text":"Modern-Day Data Science"},
  {"content_id":4,"original_text":"web-based FRONT-end development","converted_text":"Web-Based Front-End Development"}
]

Explanation

Each letter segment is independently normalized while spaces and punctuation stay unchanged.

Example 2

Input

user_content:
| content_id | content_text                          |
|------------|---------------------------------------|
| 10         |   mIXed   SPACING  and-HYPHEN   text  |
| 11         | alpha@@BETA//gamma^^delta             |

Output

[
  {"content_id":10,"original_text":"  mIXed   SPACING  and-HYPHEN   text  ","converted_text":"  Mixed   Spacing  And-Hyphen   Text  "},
  {"content_id":11,"original_text":"alpha@@BETA//gamma^^delta","converted_text":"Alpha@@Beta//Gamma^^Delta"}
]

Explanation

Repeated separators do not collapse. Each separator resets segment capitalization for the next letter.

Example 3

Input

user_content:
| content_id | content_text                    |
|------------|---------------------------------|
| 21         | pre,POST^merge\rollback       |
| 22         | one-two/THREE@four,five         |

Output

[
  {"content_id":21,"original_text":"pre,POST^merge\rollback","converted_text":"Pre,Post^Merge\Rollback"},
  {"content_id":22,"original_text":"one-two/THREE@four,five","converted_text":"One-Two/Three@Four,Five"}
]

Explanation

Commas, slashes, carets, and backslashes are preserved while still splitting letter segments.

Constraints

1 <= number of rows in user_content <= 200000
content_id is unique
1 <= content_id <= 1000000000
1 <= length(content_text) <= 400
content_text contains only English letters and these separators: space, hyphen, @, slash, caret, comma, backslash
All separators must remain exactly unchanged in output
Segment casing is applied independently between separators
SQL submissions run on SQLite 3.27.2
Pandas submissions must return columns [content_id, original_text, converted_text]