renjietang / magicktable Goto Github PK

Right now tiling & retrieving of tiles is done based on file name. If the same file is uploaded, it gets saved as name.csv and name_1.csv which can cause issues.

Ideal scenario: save the file in a db and generate some unique ID we will use inside our application rather than file name

Need to clean up the UI for the form.

• Add links to see uploaded tables as map
• Add links to download tables as csv

https://stackoverflow.com/questions/38511373/change-the-color-of-text-within-a-pandas-dataframe-html-table-python-using-style

https://stackoverflow.com/questions/42629171/how-to-use-style-in-conjunction-with-the-to-html-classes-on-a-dataframe

Pandas profiling https://github.com/pandas-profiling/pandas-profiling generates HTML profile of input dataframe.

We need to either
a. display it as a separate page/tab
b. display specific parts of it next to the map

Current implementation:
first n rows are converted to tiles and the map page is shown.
other rows are converted to tiles in the background

if the user scrolls on the map and the tiles are not yet generated, a placeholder shows. leaflet.js does not automatically refresh that placeholder.

File too large error

• center column name

• capitalize only first char in column name

• column width in each chunk of table (n rows) is rendered differently. Do scan through and calc max column width.

current URL is http://127.0.0.1:8000/tiler/list
http://127.0.0.1:8000 just lists all URLs, make it open tiler list

cloudmatcher.io/browse

To fix #4 we added padding to tiles that are smaller than 256x256.
Could we do without that?

Maybe calculate the number of rows that will fit in 256x256 and break up the CSV with those row numbers.

Actual tile:

Appears as:

This is because the actual tile is <256x256 and leaflet.js expects a 256x256 image.