summaryrefslogtreecommitdiff
path: root/TODO
diff options
context:
space:
mode:
authorNick White <hg@njw.me.uk>2011-08-01 18:26:56 +0100
committerNick White <hg@njw.me.uk>2011-08-01 18:26:56 +0100
commit063a0de3c10d38741c939297bb19d8284757e00c (patch)
treeff7b00c71dfc1bdd08b4fd1c143aaef5a0e12b9f /TODO
parentaccfeb090db432165ebc7e80ddce5ab673631af4 (diff)
Lots of tightening and cleanups
Diffstat (limited to 'TODO')
-rw-r--r--TODO28
1 files changed, 11 insertions, 17 deletions
diff --git a/TODO b/TODO
index 43cb56a..a1acd78 100644
--- a/TODO
+++ b/TODO
@@ -1,12 +1,6 @@
-Note: looks like google allows around 3 page requests per cookie session, and about 40 per ip per [some time period]. If I knew the time period, and once stdin retry is working, could make a script that gets all it can, gets a list of failures, waits, then tries failures, etc. Note these would also have to stop at some point; some pages just aren't available
+list all binaries in readme and what they do
-make sure i'm checking all lib calls that could fail
-
-make sure all arrays are used within bounds
-
-strace to check paths taken are sensible
-
-use defined constants rather than e.g. 1024
+# other utils
getgbooktxt (different program as it gets from html pages, which getgbook doesn't any more)
@@ -14,9 +8,7 @@ getabook
getbnbook
-openlibrary.org?
-
-# once it is basically working #
+# other todos
try supporting 3xx in get, if it can be done in a few lines
by getting Location line, freeing buf, and returning a new
@@ -24,14 +16,16 @@ try supporting 3xx in get, if it can be done in a few lines
add https support to get
-to be fast and efficient it's best to crank through all the json 1st, filling in an array of page structs as we go
- this requires slightly fuller json support
- could consider making a json reading module, ala confoo, to make ad-hoc memory structures from json
-
-write helper scripts like trymissing
-
write some little tests
+## getgbook
+
have file extension be determined by file type, rather than assuming png
think about whether default functionality should be dl all, rather than -a
+
+to be fast and efficient it's best to crank through all the json 1st, filling in an array of page structs as we go
+ this requires slightly fuller json support
+ could consider making a json reading module, ala confoo, to make ad-hoc memory structures from json
+
+Note: looks like google allows around 3 page requests per cookie session, and about 40 per ip per [some time period]. If I knew the time period, could make a script that gets all it can, gets a list of failures, waits, then tries failures, etc. Note these would also have to stop at some point; some pages just aren't available