Skip to content

Commit

Permalink
Add overview to README
Browse files Browse the repository at this point in the history
  • Loading branch information
AlienKevin committed Oct 10, 2023
1 parent 3709240 commit 38ead6c
Show file tree
Hide file tree
Showing 2 changed files with 17 additions and 5 deletions.
20 changes: 16 additions & 4 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,19 @@
# Configure Docker Desktop
If you are using Docker Desktop, change the allocated disk space to **200GB** in the Docker preferences (Preferences -> Resources -> Disk).
# Overview
In this artifact, we included the implementation of our web automation program synthesizer called Arborist. We will evaluate its efficiency in synthesizing programs for challenging web automation tasks and compare it with state-of-the-art techniques. In the following, we first present benchmarks we used in our evaluation, then show instructions to reproduce our experiment.

# Run main experiment
# Benchmarks
We collected 131 real-world benchmarks on web-automation tasks, with each benchmark stored in a folder with name WxTx. Each benchmark contains the following files/folder:
* *.program: the ground truth web automation program.
* script.py: the selenium program used to record the action trace and doms of the simulated user interactions on the doms.
* trace.json: the json file contains a list of actions for completing the benchmark task. Each action contains its type, corresponding selectors, and the data if applicable.
* doms/: the folder contain all website dom indexed from the starting website dom before applying the first action to the last website dom after applying the last action.

# Evaluation Instructions

## Configure Docker Desktop
If you are using Docker Desktop, change the allocated disk space to **200GB** in the Docker preferences (Preferences -> Resources -> Disk). Docker won't necessarily use all 200GB but this amount is required to prevent out of memory errors.

## Run main experiment
1. Pull and run the docker container for arborist:
```
docker run -ti alienkevin/arborist
Expand All @@ -10,7 +22,7 @@ docker run -ti alienkevin/arborist
2. Unzip test folder
```
mkdir test
tar -I pigz -xf test.tar.gz --directory tests
tar -I pigz -xf tests.tar.gz --directory tests
```
Ignore warnings like below:
```
Expand Down
2 changes: 1 addition & 1 deletion build.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
# Zip benchmarks for Docker
```
tar -c --use-compress-program=pigz -f test.tar.gz tests
tar -c --use-compress-program=pigz -f tests.tar.gz tests/benchmarks/W8T3 tests/benchmarks/W308T1 tests/benchmarks/W48T1 tests/benchmarks/W156T1 tests/benchmarks/W112T1 tests/benchmarks/W40T1 tests/benchmarks/W56T1 tests/benchmarks/W80T1 tests/benchmarks/W307T2 tests/benchmarks/W295T1 tests/benchmarks/W239T2 tests/benchmarks/W204T1 tests/benchmarks/W110T2 tests/benchmarks/W218T1 tests/benchmarks/W239T1 tests/benchmarks/W307T1 tests/benchmarks/W6T1 tests/benchmarks/W261T2 tests/benchmarks/W253T1 tests/benchmarks/W169T1 tests/benchmarks/W228T3 tests/benchmarks/W228T2 tests/benchmarks/W218T2 tests/benchmarks/W144T1 tests/benchmarks/W120T1 tests/benchmarks/W228T1 tests/benchmarks/W91T2 tests/benchmarks/W268T1 tests/benchmarks/W49T1 tests/benchmarks/W250T1 tests/benchmarks/W303T1 tests/benchmarks/W110T3 tests/benchmarks/W302T1 tests/benchmarks/W8T2 tests/benchmarks/W8T1 tests/benchmarks/W265T2 tests/benchmarks/W265T1 tests/benchmarks/W228T4 tests/benchmarks/W274T1 tests/benchmarks/W87T1 tests/benchmarks/W133T1 tests/benchmarks/W1T1 tests/benchmarks/W138T1 tests/benchmarks/W154T1 tests/benchmarks/W232T2 tests/benchmarks/W240T1 tests/benchmarks/W99T1 tests/benchmarks/W1T2 tests/benchmarks/W284T1 tests/benchmarks/W296T1 tests/benchmarks/W162T1 tests/benchmarks/W309T1 tests/benchmarks/W178T1 tests/benchmarks/W33T1 tests/benchmarks/W263T2 tests/benchmarks/W278T1 tests/benchmarks/W134T1 tests/benchmarks/W158T2 tests/benchmarks/W189T1 tests/benchmarks/W49T2 tests/benchmarks/W34T2 tests/benchmarks/W252T1 tests/benchmarks/W158T1 tests/benchmarks/W252T2 tests/benchmarks/W25T1 tests/benchmarks/W287T2 tests/benchmarks/W237T1 tests/benchmarks/W87T2 tests/benchmarks/W34T3 tests/benchmarks/W69T1 tests/benchmarks/W232T1 tests/benchmarks/W77T1 tests/benchmarks/W34T1 tests/benchmarks/W214T1 tests/benchmarks/W226T1 tests/benchmarks/W124T1 tests/benchmarks/W287T1 tests/benchmarks/W74T1 tests/benchmarks/W139T1 tests/benchmarks/W115T1 tests/benchmarks/W304T2 tests/benchmarks/W262T1 tests/benchmarks/W173T1 tests/benchmarks/W74T2 tests/benchmarks/W304T1 tests/benchmarks/W1T3 tests/benchmarks/W164T1 tests/benchmarks/W223T1 tests/benchmarks/W141T1 tests/benchmarks/W305T1 tests/benchmarks/W148T1 tests/benchmarks/W146T1 tests/benchmarks/W233T1 tests/benchmarks/W81T1 tests/benchmarks/W125T1 tests/benchmarks/W146T2 tests/benchmarks/W263T1 tests/benchmarks/W285T1 tests/benchmarks/W190T1 tests/benchmarks/W213T1 tests/benchmarks/W157T1 tests/benchmarks/W157T2 tests/benchmarks/W276T1 tests/benchmarks/W53T1 tests/benchmarks/W127T1 tests/benchmarks/W254T1 tests/benchmarks/W88T1 tests/benchmarks/W69T2 tests/benchmarks/W54T2 tests/benchmarks/W205T1 tests/benchmarks/W91T3 tests/benchmarks/W51T2 tests/benchmarks/W18T1 tests/benchmarks/W188T1 tests/benchmarks/W14T1 tests/benchmarks/W46T1 tests/benchmarks/W52T1 tests/benchmarks/W111T2 tests/benchmarks/W7T2 tests/benchmarks/W9T1 tests/benchmarks/W50T1 tests/benchmarks/W78T2 tests/benchmarks/W111T1 tests/benchmarks/W176T1 tests/benchmarks/W177T1 tests/benchmarks/W149T1 tests/benchmarks/W3T1 tests/benchmarks/W58T1 tests/benchmarks/W149T2 tests/benchmarks/W238T1 tests/benchmarks/W38T1
```

# Building Docker for Multi-Platform
Expand Down

0 comments on commit 38ead6c

Please sign in to comment.