文件列表:

TODO.md (493, 2019-09-11)
package-lock.json (525039, 2019-09-11)
package.json (785, 2019-09-11)
public (0, 2019-09-11)
public\data (0, 2019-09-11)
public\data\2019 - April.json (1076935, 2019-09-11)
public\data\2019 - August.json (916672, 2019-09-11)
public\data\2019 - February.json (1051658, 2019-09-11)
public\data\2019 - January.json (1135248, 2019-09-11)
public\data\2019 - July.json (1093174, 2019-09-11)
public\data\2019 - June.json (973465, 2019-09-11)
public\data\2019 - March.json (1110927, 2019-09-11)
public\data\2019 - May.json (1183011, 2019-09-11)
public\data\2019 - September.json (785966, 2019-09-11)
public\favicon.ico (3870, 2019-09-11)
public\index.html (1967, 2019-09-11)
public\manifest.json (306, 2019-09-11)
screenshot.png (903670, 2019-09-11)
src (0, 2019-09-11)
src\JobFilters.js (1555, 2019-09-11)
src\JobGraphs.js (2465, 2019-09-11)
src\JobResults.js (1574, 2019-09-11)
src\JobSources.js (1603, 2019-09-11)
src\JobStats.js (1968, 2019-09-11)
src\StatusBar.js (236, 2019-09-11)
src\data (0, 2019-09-11)
src\data\JobsData.js (9768, 2019-09-11)
src\index.css (1988, 2019-09-11)
src\index.js (467, 2019-09-11)
src\lib (0, 2019-09-11)
src\lib\JobDownloader.mjs (4702, 2019-09-11)
src\lib\getJSON.js (1200, 2019-09-11)
src\logo.svg (2671, 2019-09-11)
src\scripts (0, 2019-09-11)
src\scripts\get_hn_jobs.mjs (909, 2019-09-11)
src\serviceWorker.js (4951, 2019-09-11)
yarn.lock (328868, 2019-09-11)
... ...

 In order to run this, clone / download the repository. CD into the directory and run: > npm install > yarn start If your web browser doesn't already open, then go to http://localhost:3000/ and the app will begin to load. It takes some time initially, but then you should be able to filter with ease. -------------------------------------- 8/2/2019: - Productive day today. Added some style to the project, got deleting of filters working, added July and August '19 jobs data and fixed a handful of minor bugs along the way. -------------------------------------- 7/30/2019: - Made the graphs work yesterday but it also gave me some ideas on how to make the whole thing more useful - Venn diagrams, while pretty are probably not that useful. Also, I'll just create a "todo.md" file to track all todos / future featuers. -------------------------------------- 7/24/2019: - We now have enough data to start drawing stuff. Next step for data: - Jobs from multiple years. - We will want to lazyload this based on demand. -------------------------------------- 7/22/2019: Okay, so now we have enough to make our data be more visual using D3.js. What'd be great on the current dataset: - Venn diagram: - See all jobs. - See each filter. - See the intersection of filters. - Be able to click on different parts to drill down on the data (check / uncheck filters). - Checking / unchecking the checklist will transition the ven diagram to reflect the new state. -------------------------------------- 7/18/2019: The new direction is good. We're going to download all the job posts and store them as static json files on server. The client will then load up the json files lazily and apply users' filters on them as well as prepare any analytics. First load is a bit slow right now, which will probably become an issue soon enough. -------------------------------------- 7/16/2019: Problem: Execute a function after a function that runs on an array has completed on the last element of the array... such as downloading data from the last URL. While you're at it, only begin processing the next item, when the previous item has completed. My new favorite solution (thank you, Elixir): Use recursion. Call processing function on the whole array. In the function, pop an element off, do the processing. Recursively call the same function with whats left of the array. And finally, when there are no more elements left (the passed in array is empty), either resolve the promise or run the callback funtion! This also makes it much easier to write functions that process data as promises. -------------------------------------- 7/15/2019: Problem: - If we're going to scrape all that data, then it is probably a good idea to store it on the server, so we don't keep hammering the API. - Since the amount of data can get sizeable, the frontend should probably only get enough data that it needs for displaying: - Menu items. - Current results. - Data for charts. So, what will our backend do? - Populate the database - Provide an API for: - Creating filters - Calculate results for filters - Requesting a resultset of enabled filters This leaves the front-end to: - Display results - Display sources - Display filters - Allow creation of new filters, which the backend will then sift through the data to provide an array if IDs to the front-end. - Calculate results from enabled filters and sources, request results for missing jobs. ... or maybe not. - Grab all historical posts from HN. - Add them to JSON files on the server. - Lazy load them as needed.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：