Systems and methods for a search tool of code snippets
The present application generally relates to search engines, and more specifically to systems and methods for a search tool for code snippets. Embodiment described herein provide a customized code search system that generates code search results from customized data sources, extract code snippets from the code search results, and presents the code snippets via a user interface. In one embodiment, the search system adopts a machine learning module to generate and highlight search results from different data sources that include code examples, e.g., in a programming language. To improve search efficiency, in response to a code search query, the search system may extract code snippets from search results from relevant sources in a user interface element, such as user selectable panels.
1. A processor-implemented method of performing real-time code search in an integrated development environment (IDE), the method comprising:
monitoring, via a client component with an IDE implemented on a user device, user activities relating to a code segment;
determining, by a neural network based prediction model implemented on one or more hardware processors, a code search query based at least in part on a first portion of the code segment and the monitored user activities;
transmitting, via a first search application programming interface (API) over a network, a first search input customized from the code search query to a first data source that has been pre-categorized for code search;
receiving, from the first search API, a first search result including a first link to a first webpage containing a first code snippet in response to the code search query;
extracting the first code snippet from the first webpage following the first link; and
causing a display of a first user-engageable widget displaying the first code snippet at a user interface within the IDE.
2. The method of claim 1 , wherein the code search query is determined further based on a user activity event indicative of a user intention for the code search.
3. The method of claim 1 , wherein the code search query is determined further based on identifying an error within a first portion of the code segment.
4. The method of claim 1 , further comprising:
transmitting, via a second search application programming interface (API) and over the network, a second search input customized from the code search query to a second data source that has been pre-categorized for code search;
receiving, from the second search API, a second search result including a second link to a second webpage containing a second code snippet in response to the code search query;
extracting the second code snippet from the second webpage following the second link;
ranking, by a ranker model, the first search result and the second search result; and
causing a display of the first user-engageable widget displaying the first code snippet and a second user-engageable widget displaying the second code snippet according to the ranking at a user interface within the IDE.
5. The method of claim 1 , wherein the first user-engageable widget includes a visual element indicating the first data source.
6. The method of claim 1 , wherein the monitored user activities are transmitted to a search server, and wherein the neural network based prediction model is implemented at the search server, and wherein the method further comprises:
transmitting the first code snippet to the user device thereby causing the display of a first user-engageable widget.
7. The method of claim 1 , wherein the neural network based prediction model is implemented by the client component at the user device.
8. The method of claim 1 , wherein the client component receives a user selection of the first code snippet, thereby causing automatic integration of the first code snippet into the code segment upon the user selection.
9. The method of claim 8 , wherein the automatic integration of the first code snippet comprises a replacement of a sub-portion of the code segment that has been identified as relating to a code error with the first code snippet.
10. A system for performing real-time code search in an integrated development environment (IDE), the system comprising:
a communication interface that receives, via a client component with an IDE implemented on a user device, user activities relating to a code segment;
a physical memory storing neural network based models and a plurality of processor-executable instructions; and
one or more physical processors executing the instructions to perform operations comprising:
determining, by a neural network based prediction model implemented on one or more hardware processors, a code search query based at least in part on a first portion of the code segment and the monitored user activities;
transmitting, via a first search application programming interface (API) over a network, a first search input customized from the code search query to a first data source that has been pre-categorized for code search;
receiving, from the first search API, a first search result including a first link to a first webpage containing a first code snippet in response to the code search query;
extracting the first code snippet from the first webpage following the first link; and
causing a display of a first user-engageable widget displaying the first code snippet at a user interface within the IDE.
11. The system of claim 10 , wherein the code search query is determined further based on a user activity event indicative of a user intention for the code search.
12. The system of claim 10 , wherein the code search query is determined further based on identifying an error within a first portion of the code segment.
13. The system of claim 10 , wherein the operations further comprise:
transmitting, via a second search application programming interface (API) and over the network, a second search input customized from the code search query to a second data source that has been pre-categorized for code search;
receiving, from the second search API, a second search result including a second link to a second webpage containing a second code snippet in response to the code search query;
extracting the second code snippet from the second webpage following the second link;
ranking, by a ranker model, the first search result and the second search result; and
causing a display of the first user-engageable widget displaying the first code snippet and a second user-engageable widget displaying the second code snippet according to the ranking at a user interface within the IDE.
14. The system of claim 10 , wherein the first user-engageable widget includes a visual element indicating the first data source.
15. The system of claim 10 , wherein the client component receives a user selection of the first code snippet, thereby causing automatic integration of the first code snippet into the code segment upon the user selection.
16. The system of claim 15 , wherein the automatic integration of the first code snippet comprises a replacement of a sub-portion of the code segment that has been identified as relating to a code error with the first code snippet.
17. A processor-readable non-transitory storage medium storing a plurality of processor-executable instructions for performing real-time code search in an integrated development environment (IDE), the instructions being executed by one or more processors to perform operations comprising:
monitoring, via a client component with an IDE implemented on a user device, user activities relating to a code segment;
determining, by a neural network based prediction model implemented on one or more hardware processors, a code search query based at least in part on a first portion of the code segment and the monitored user activities;
transmitting, via a first search application programming interface (API) over a network, a first search input customized from the code search query to a first data source that has been pre-categorized for code search;
receiving, from the first search API, a first search result including a first link to a first webpage containing a first code snippet in response to the code search query;
extracting the first code snippet from the first webpage following the first link; and
causing a display of a first user-engageable widget displaying the first code snippet at a user interface within the IDE.
18. The processor-readable non-transitory storage medium of claim 17 , wherein the code search query is determined further based on a user activity event indicative of a user intention for the code search.
19. The processor-readable non-transitory storage medium of claim 17 , wherein the code search query is determined further based on identifying an error within a first portion of the code segment.
20. The processor-readable non-transitory storage medium of claim 1 , wherein the operations further comprise:
transmitting, via a second search application programming interface (API) and over the network, a second search input customized from the code search query to a second data source that has been pre-categorized for code search;
receiving, from the second search API, a second search result including a second link to a second webpage containing a second code snippet in response to the code search query;
extracting the second code snippet from the second webpage following the second link;
ranking, by a ranker model, the first search result and the second search result; and
causing a display of the first user-engageable widget displaying the first code snippet and a second user-engageable widget displaying the second code snippet according to the ranking at a user interface within the IDE.