Re: [libvirt PATCHv2 10/15] nss: convert findLeases to use json-c

On Thu, Sep 05, 2024 at 15:49:37 +0200, Ján Tomko wrote: > While the parsing is still done by 1K buffers, the results > are no longer filtered during the parsing, but the whole JSON > has to live in memory at once, which was also the case before > the NSS plugin dropped its dependency on libvirt_util. > > Also, the new parser might be more forgiving of missing elements. > > Signed-off-by: Ján Tomko <jtomko(a)redhat.com&gt; > --- > tools/nss/libvirt_nss_leases.c | 370 +++++++++++---------------------- > 1 file changed, 119 insertions(+), 251 deletions(-)

> +/** > + * findLeaseInJSON > + * > + * @jobj: the json object containing the leases > + * @name: the requested hostname (optional if a MAC address is present) > + * @macs: the array of MAC addresses we're matching (optional if we have a hostname) > + * @nmacs: the size of the MAC array > + * @af: the requested address family > + * @now: current time (to eliminate expired leases) > + * @addrs: the returned matching addresses > + * @naddrs: size of the returned array > + * @found: whether a match was found > + * > + * Returns 0 even if nothing was found > + * -1 on error > + */ > static int > +findLeaseInJSON(json_object *jobj, > + const char *name, > + char **macs, > + size_t nmacs, > + int af, > + time_t now, > + leaseAddress **addrs, > + size_t *naddrs, > + bool *found) > { > size_t i; > + int len; > > + if (!json_object_is_type(jobj, json_type_array)) { > + ERROR("parsed JSON does not contain the leases array"); > + return -1; > } > > + len = json_object_array_length(jobj); > + for (i = 0; i < len; i++) { > + json_object *lease = NULL; > + json_object *expiry = NULL; > + json_object *ipobj = NULL; > + unsigned long long expiryTime; > + const char *ipaddr; > + > + lease = json_object_array_get_idx(jobj, i); > + > + if (macs) { > + const char *macAddr; > + bool match = false; > + json_object *val; > + size_t j; > + > + val = json_object_object_get(lease, "mac-address"); > + if (!val) > + continue; > + > + macAddr = json_object_get_string(val); > + if (!macAddr) > + continue; > + > + for (j = 0; j < nmacs; j++) { > + if (!strcmp(macs[j], macAddr)) { strcmp(..) != 0

> + match = true; > + break; > + } > + } > + if (!match) > @@ -389,51 +271,37 @@ findLeases(const char *file, > goto cleanup; > } > > + tok = json_tokener_new(); > + json_tokener_set_flags(tok, jsonflags); > > + do { > + rv = read(fd, line, sizeof(line) - 1); -1 should not be needed here > if (rv < 0) > goto cleanup; > if (rv == 0) > break; So if you get an incomplete JSON document of exactly sizeof(len) bytes, this will go through the first loop adn thend break out ... > nreadTotal += rv; > > + jobj = json_tokener_parse_ex(tok, line, rv); > + jerr = json_tokener_get_error(tok); > + } while (jerr == json_tokener_continue); >

> + if (nreadTotal > 0 && jerr != json_tokener_success) { ... this condition will match as we read something and 'jerr' is still 'json_tokener_continue' > + ERROR("Cannot parse %s: %s", file, json_tokener_error_desc(jerr)); So this error will read: Cannot parse /path/to/file: continue