instagrab 0.1.0

Scrapes Instagram profile metadata by attaching to a real Chrome over CDP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

# instagrab config

# CDP endpoint of the long-lived Chrome service.
browser_url = "http://127.0.0.1:9222"

# JSONL file. Each run appends one line per username plus any alert lines.
output_path = "/var/lib/instagrab/runs.jsonl"

# Per-run sleep between profiles, in seconds (uniform random in [min, max]).
jitter_min_secs = 30
jitter_max_secs = 75

# Per-username navigation/wait budget.
nav_timeout_sec = 20

# Only keep recent_posts whose taken_at_timestamp falls within the last N days.
# 0 disables the filter. With time_window_days > 0, instagrab also paginates
# (scrolls the profile grid) past the initial 12 posts until the oldest
# captured post is older than the window.
time_window_days = 7

# Safety cap on pagination scrolls per profile. Each iteration is ~2.5s.
max_scrolls = 8

# If set, download each kept post's display_url to <images_dir>/<username>/<shortcode>.jpg.
# Idempotent: existing files are kept. Empty string disables image download.
images_dir = "./images"

# Profiles to scrape. Leading @ is allowed and stripped.
usernames = ["inch.press"]