How Manual Searches Work v2.5

➲ Manual Queries Overview

Before diving into manual queries, it helps to know how Diskover reads the text you type into the search bar. File names are often messy—different teams, tools, and workflows create their own styles, so it’s common to see inconsistent spelling, odd separators, abbreviations, and creative numbering.

Diskover follows predictable rules so you can understand why certain results appear—or why others don’t.

Your industry might differ from the examples shown, but the search logic is universal—so we highlight only the core concepts you’ll use most.

✏️ The rules addressed in this section do not apply to Regex (regular expression) searches, which use their own syntax and are intended for advanced users.

➲ Many Ways to Search

Diskover gives you multiple ways to search—you can use:
• Built-in tools.
• Manual queries, as described in this section.
• A combination of both built-in tools and manual queries.

➲ Golden Rules of Searching

Searching is an iterative process!

• Start broad, then narrow down: Use tools like wildcards to make sure you’re not unintentionally missing files or folders—missing expected items is one of the most common issues when searching.

• Add one search criterion at a time: Review and validate your results after each addition to confirm you’re heading in the right direction.

• Don’t assume the file is named the way you expect: Names vary (misspellings, abbreviations, underscores, “v8” vs “ver08”), so start with broader search terms and refine from there.

• Try removing criteria if your results feel “too empty.”: If you’re getting no results or far fewer than expected, simplify the query to confirm whether the issue is the search or the criteria.

• Switch methods: If a manual query isn’t giving you what you need, try a built-in search tool, for example, and then edit the query.

➲ Search Rules Based on Elasticsearch

Diskover uses Elasticsearch to store metadata and power all searches. Because of this, Diskover’s search syntax follows Elasticsearch rules and behavior.

Learn more about Elasticsearch String Query →

➲ Naming Conventions

Many organizations deal with inconsistent or messy naming conventions—misspellings, abbreviations, multiple formats, or teams naming the same thing differently.

Here are just a few examples:

In situations like this, it helps to use tools like wildcards for broader, more flexible searches.

➲ Case Sensitivity

Searches are usually case-insensitive, meaning uppercase and lowercase behave the same.
There are a few exceptions when searching on:

• Timestamps

• Field names

➲ Tokenizers | How Words and Numbers Are Broken Apart

When you type a word or number, Diskover looks for separated characters.

To figure out what’s separated, Diskover uses a tokenizer that splits text, such as:

• spaces

• underscores

• hyphens

• periods

• punctuation

• capital letters, aka CamelCase (text like AlbertSimulationTest is treated as Albert + Simulation + Test)

If a word or number isn’t separated, Diskover may not match/find it—unless you use wildcards.

Tokenization | Examples with Letters

🔎 Searching for albert

Tokenization | Examples with Numbers

🔎 Searching for 2025

➲ What is a Key-Value Pair

In Diskover, key-value pairs are used in many places, including search queries, reports, workflows, and building datasets for AI/BI pipelines.

• A key identifies the field you want to search.

• A value is the variable you are looking for.

Together, they create a simple way to define search criteria.

Key-Value Pair Examples

The examples below may look complex, but you can use built-in search tools to generate most queries like these automatically.

When combining multiple criteria, you’ll use operators such as AND, NOT, OR, and parentheses for grouping.