nx_page_fetch_tmp.nx source
↩ module page · 43 lines · 2053 B
1// nx_page_fetch_tmp.nx -- fetch ONE page -> strip to text -> /tmp/page.txt.
2// Transient research-time input (NOT a baked corpus). Imports only the fetcher
3// + html stripper (disjoint from the research-organ import tree) so the two
4// never collide at compile time. expect_exit: 0.
5
6import "nx_syscalls.nx"
7import "nx_x509_trust_store.nx"
8import "nx_trust_store_load_from_certdata.nx"
9import "nx_https_fetch_follow.nx"
10import "nx_html_to_text.nx"
11const K_MAGIC_4194304: i64 = 4194304
12const K_MAGIC_8388608: i64 = 8388608
13
14func w(s: *u8) -> i64 { var n: i64 = 0; while s[n] != (0 as u8) { n = n + 1 } sys_write(1, s, n); return 0 }
15func wn(v: i64) -> i64 {
16 let b: *u8 = sys_mmap(28); var m: i64 = v
17 if m < 0 { m = 0 - m; sys_write(1, "-" as *u8, 1) }
18 let t: *u8 = sys_mmap(28); var k: i64 = 0
19 if m == 0 { t[0] = 48 as u8; k = 1 }
20 while m > 0 { t[k] = (48 + (m % 10)) as u8; m = m / 10; k = k + 1 }
21 var i: i64 = 0; while i < k { b[i] = t[k - 1 - i]; i = i + 1 }
22 sys_write(1, b, k); return 0
23}
24
25func main() -> i64 {
26 let r: i64 = nx_trust_store_load_from_certdata("/tmp/mozilla_certdata.txt" as *u8, 300, K_MAGIC_4194304)
27 if r <= 0 { w("STORE-FAIL\n" as *u8); return 1 }
28 let store: *TrustStore = r as *TrustStore
29 let raw: *u8 = sys_mmap(K_MAGIC_8388608)
30 let txt: *u8 = sys_mmap(K_MAGIC_8388608)
31 let st: *i64 = (sys_mmap(8)) as *i64
32 let n: i64 = nx_https_fetch_follow("https://en.wikipedia.org/wiki/Game_engine" as *u8, store, raw, K_MAGIC_8388608, 6, st)
33 w("fetch status=" as *u8); wn(st[0]); w(" raw=" as *u8); wn(n); w("\n" as *u8)
34 if n <= 0 { w("FETCH-FAIL\n" as *u8); return 1 }
35 let tn: i64 = nx_html_to_text(raw, n, txt, K_MAGIC_8388608)
36 let fd: i64 = sys_openat_wr("/tmp/page.txt" as *u8, 420)
37 if fd < 0 { w("OPEN-FAIL\n" as *u8); return 1 }
38 var wr: i64 = 0
39 while wr < tn { let c: i64 = sys_write(fd, txt + wr, tn - wr); if c <= 0 { break } wr = wr + c }
40 sys_close(fd)
41 w("wrote text_bytes=" as *u8); wn(wr); w(" -> /tmp/page.txt\n" as *u8)
42 return 0
43}