radcli_avp_concat_str(3)

📦 version 2.0.1, in arch rolling

💅 radcli_avp_concat_str — from the radcli manual, section 3

okay so bestie, buckle up, because this one is literally the "grab every copy of an attribute and smoosh them into one string" function, and it ATE. it's radcli_avp_concat_str(3) and it's about to make your reply-parsing life so much easier 🫶

✨ SYNOPSIS ✨

#include <radcli/>

int radcli_avp_concat_str(
    char                   *buf,
    size_t                  buflen,
    const radcli_avp_list  *list,
    const radcli_attr_def  *def,
    const char             *sep
);

🌸 DESCRIPTION 🌸

so basically this generalizes the one little convenience thing that legacy radcli.h's rc_aaa() used to fold right into its own call signature — you know, its msg parameter, the one that "will contain the concatenation of any PW_REPLY_MESSAGE received" (lib/buildreq.c) 💫 — except now it works for any attribute, not just Reply-Message. because the pattern "walk every occurrence of one attribute in a reply, join as text" was never actually specific to Reply-Message, and everyone knew it. so if you're a caller who just wants rc_aaa()'s old convenience? you no longer have to hand-roll the whole radcli_avp_get()-in-a-loop thing that this function replaces. no cap, that loop can retire 💅

if there's no occurrence of def present in list, buf gets set to an empty string and 0 is returned — and that is not a failure, okay? it's the exact same energy as rc_aaa()'s msg starting out as '' and just... staying that way when no Reply-Message ever showed up. iconic, honestly. and an attribute whose value contains an embedded NUL byte, or (for a RADCLI_TYPE_TEXT attribute) invalid UTF-8, gets skipped — that's radcli_avp_get_cstr()'s policy — and it's also not treated as a failure. no drama, just vibes 🌈

it follows snprintf()'s buffer-sizing contract exactly, and when I say exactly I mean it:

  • buf may be NULL and/or buflen may be 0 so you can size a buffer before you allocate one (in that case nothing gets written, periodt)
  • the return value is always the number of bytes the joined result would occupy, whether or not it actually fit
  • so a return >= buflen means the result was truncated, and buf (if it's non-NULL and buflen > 0) then holds only a valid NUL-terminated prefix of whatever fits — never garbage, never unterminated 🙈

🎀 PARAMETERS 🎀

  • buf — the destination buffer, or NULL if you only want to compute the needed size. it's always left NUL-terminated on return (truncation included, holding whatever fit) whenever buf is non-NULL and buflen > 0 🦋
  • buflen — size of buf, in bytes. may be 0.
  • list — the list you're searching. your scavenger hunt ✨
  • def — the attribute you want to concatenate occurrences of.
  • sep — the separator inserted between occurrences. NULL or "" for none. (yes, both work, live your truth)

👑 RETURN VALUE 👑

the number of bytes the joined result occupies (excluding the NUL terminator), whether or not it fit in buf — or -1 if def is NULL. that's it, that's the deal 🔥

💜 SEE ALSO 💜

your reading list, bestie, and yes you're gonna read all of them:

friendly reminder from your local Unix bestie: buf may be NULL and buflen may be 0 — that's the size-only mode, nothing is written, don't panic 💀

okay that's the whole page, mwah 💋 this came out of the radcli source, dated 2026-10-10, and the page itself is radcli_avp_concat_str(3) — go forth and concatenate, you absolute main character 🌟

📦 where this page comes from