@dewen_li/migration-tool NPM

Site Migration tool

The goal of this tool is to take an existing site (independent of its technology) and convert it static flattened html site. A flattened html site is non-editable site that can be served with no databases with minimal platform requirements. The tool automates this flattening to reduce the cost of migration of these sites off other technology.

The tool is given URL that will crawl the existing site finding all pages, generates the html, pulls the assets of the sites and outputs the static html site. The tool is provided as part of the BodilessJS project and is incorporated into migration process. The output works in coordination with the BodilessJS Starter Kit to use the same tooling/infrastructure.

Features

Crawls the website and gets a list of website pages
Scrapes website's page html
Scrapes website's assets:
1. Scripts
2. Css styles
3. Images
Scrapes website's meta tags
Provide html transformation rules
- rules to remove html elements from dom
- rules to replace html elements with html snippet
Generate BodilessJS pages
Converts html into jsx (jsx is a React extension that allows us to write JavaScript that looks like HTML.)
Breaks monolithic html into pieces

Getting Started

Prerequisites

Node.js version >= 10
npm

Installation & Run

Clone this repository: https://github.com/johnsonandjohnson/bodiless-js.git
Execute the command to install the necessary packages

npm run setup

npm run fresh

to include cleaning step before the setup.

Prepare the migration-settings.json. All settings can be found in Configuration section
Execute the command to flatten the site

npm run migrate

Configuration

Using the requirements of the flattened site, prepare a build plan by adjusting migration-settings.json

The following options control how the the site will be flattened.

Options:

url
- Description: Url of website that should be flattened
- Accepted Formats:: Preface the url with http:// or https:// syntax
- Examples:: "http://pariet10.ru/" or "https://pariet10.ru/"
- Restrictions:: None
steps
- Description: Specify a list of steps that should be executed by the tool
- Recommendation: Specify all steps and set as true.

setup

		*  **Description**: Enable/disable cloning and setting up BodilessJS app locally

		*  **Accepted Formats:**: "true" or "false"

	*  `scrape`

		*  **Description**: Enable/disable site scraping and BodilessJS pages generation

		*  **Accepted Formats:**: "true" or "false"

	*  `build`

		*  **Description**: Enable/disable building of static site

		*  **Accepted Formats:**: "true" or "false"

	*  `serve`

		*  **Description**: Enable/disable serving of static site

		*  **Accepted Formats:**: "true" or "false"

crawler
- Description: Specify configuration for the crawler
- maxDepth
  - Description: Maximum depth for the crawler to follow links automatically
  - Accepted Formats:: Number > 0
  - Recommendation:: 100, unless there is specific reason to limit. A higher number will allow tool to crawl the entire site.
- maxConcurrency
  - Description: Maximum number of pages to open concurrently
  - Accepted Formats:: Number > 0
  - Recommendation:: Recommend to use 1 to not impact production sites
- ignoreRobotsTxt
  - Description: Whether robots.txt should be ignored and whether urls disallowed in robots.txt should be scraped, default to false.
  - Accepted Formats: Boolean (true or false)
  - Recommendation: Recommend to use default value when the whole website is scraped, set this option to true once a disallowed page should be scraped individually.
htmltojsx
- Description: Enable/disable transforming html to jsx
- Accepted Formats:: "true" or "false"
- Recommendation:: "true"

transformers
- Description: Specify rules that should be applied to the scraped page html
rule
- Description: The following rules allow you to manipulate the output and either remove or replace html selector components.
- replace
  - Description: Replace each element in the set of matched elements with the provided new content and return the set of elements that was removed.
- replaceString
  - Description: Replace string (or regex pattern) in the source html code before parsing.
- tojsx
  - Description: Extract matched elements into React components as separate modules
  - Note: this option is not implemented yet. If you want to convert html into jsx please follow "Convert html into jsx" section
  - Specific configuration parameters for each rule type:
selector
- Description: Selector for the element(s) that should be processed
- Accepted Formats:: string
replacement
- Description: New html content in replace mode. Name of React component in tojsx mode
- Accepted Formats:: string
- Restrictions:: Escape special characters, such as " with \"
disableTailwind
- Description: Disables site tailwind theme. The site tailwind theme is disabled by default, set this variable to false to enable site tailwind theme. You may want to enable the tailwind theme if the migrated site will add new bodiless JS components. Note that enabling the tailwind theme in some cases may interfere with the migrated site's styling.
- Accepted Formats: true or false
- Default Value: true
allowFallbackHtml
- Description: Optional setting to push original html body into the page component file and report a message in the output when migration encounters an error from body section of html. If disabled with value false, migration skips generation of JSX on page parsing errors.
- Accepted Formats: Boolean (true or false)
- Default Value: true

context
- Description: A list of pages in which the rule should be applied.
- Accepted Formats:: Please follow minimatch syntax to compose url pattern
Examples:

{

"rule":  "replace",

"selector":  "script[src*='cdn.cookielaw.org/consent']",

"replacement":  "<script charset=\"UTF-8\" src=\"https://optanon.blob.core.windows.net/consent/086a2433-54aa-4112-8ba6-331eb1d2fda7-test.js\"></script>",

"context":  "**"

}

Rehydration (convert html elements into React Components) during Site Migration

The following process will rehydrate (or replace specific html element components with React named components.)

In the packages/bodiless-migration-tool/conf/settings.json

Specify replace Rules

selector

Description: selector for the element(s) that should be extracted into React components

component name of the React component
- Description: selector for the element(s) that should be extracted into React components

Example

View conf/settings.json

Usage

To flatten a website using the tool, run:

npm run migrate

Migration Output

Flattened & Build files/assets

The individual pages of the site can found in examples\test-site\src\data\pages

After performing of npm run build:

The output of the build site can be found in the examples\test-site\public
The assets of the site can be found in the examples\test-site\static

View Migrated Site

The flattened & generated built site will be viewable at http://0.0.0.0:9000/ OR if on windows http://localhost:9000.

Full settings.json Examples

Full examples can be found in examples/settings.

Configure no-scroll for selected anchor element.

Sites, undergoing flattening, may have foldable accordion elements that are implemented with anchor fragment. By default, GatsbyJS navigation compares the change in url and scrolls the page to the location based on the given url hash.

If this is not the desired behavior for migrated page, a user might override scrolling by configure no-scroll-settings.json and place it under [site]/src/@bodiless/gatsby-theme-bodiless/ folder. Behind the scenes, it shadows Gatsby theme packages/gatsby-theme-bodiless/src/no-scroll-settings.json configuration file.

Here's an example of no-scroll-settings.json usages:

{
  "parentSelectors": [
    ".container-classname-1",
    ".container-classname-2"
  ],
  "elementSelectors": [
    ".container-classname .fieldset .field__item > a"
  ],
  "excludeHashes": [
    "hash-to-be-excluded-from-no-scrolling"
  ]
}

Explanation of options:,

parentSelectors: Contains a list of selectors that within the selected container element, clicking anchor will disable the page scrolling.
elementSelectors: Contains a list of anchor selectors that clicking on these anchors will disable the page scrolling.
excludeHashes: Used to define a custom list of hash strings (without "#" character) to be excluded from no-scrolling feature.