Track bridge info by country first, then date

This commit is contained in:
Vecna 2024-03-23 21:40:00 -04:00
parent 3ef9c04141
commit 12e699f979
1 changed files with 90 additions and 105 deletions

View File

@ -49,11 +49,8 @@ pub struct BridgeInfo {
/// first Julian date we started collecting data on this bridge
pub first_seen: u32,
/// list of countries where the bridge is believed to be blocked
pub blocked_in: Vec<String>,
/// map of dates to data for that day
pub info_by_day: HashMap<u32, DailyBridgeInfo>,
/// map of countries to data for this bridge in that country
pub info_by_country: HashMap<String, BridgeCountryInfo>,
}
impl BridgeInfo {
@ -62,8 +59,7 @@ impl BridgeInfo {
fingerprint: fingerprint,
nickname: nickname.to_string(),
first_seen: get_date(),
blocked_in: Vec::<String>::new(),
info_by_day: HashMap::<u32, DailyBridgeInfo>::new(),
info_by_country: HashMap::<String, BridgeCountryInfo>::new(),
}
}
}
@ -76,15 +72,11 @@ impl fmt::Display for BridgeInfo {
);
str.push_str(format!("nickname: {}\n", self.nickname).as_str());
str.push_str(format!("first_seen: {}\n", self.first_seen).as_str());
str.push_str("blocked_in:");
for country in &self.blocked_in {
str.push_str(format!("\n {}", country).as_str());
}
str.push_str("info_by_day:");
for day in self.info_by_day.keys() {
str.push_str(format!("\n day: {}", day).as_str());
let daily_info = self.info_by_day.get(day).unwrap();
for line in daily_info.to_string().lines() {
str.push_str("info_by_country:");
for country in self.info_by_country.keys() {
str.push_str(format!("\n country: {}", country).as_str());
let country_info = self.info_by_country.get(country).unwrap();
for line in country_info.to_string().lines() {
str.push_str(format!("\n {}", line).as_str());
}
}
@ -99,62 +91,49 @@ pub enum BridgeInfoType {
PositiveReports,
}
/// Information about bridge reachability, gathered daily
/// Information about bridge reachability from a given country
#[derive(Serialize, Deserialize)]
pub struct DailyBridgeInfo {
pub info_by_country: BTreeMap<String, BTreeMap<BridgeInfoType, u32>>,
pub struct BridgeCountryInfo {
pub info_by_day: BTreeMap<u32, BTreeMap<BridgeInfoType, u32>>,
pub blocked: bool,
}
impl DailyBridgeInfo {
impl BridgeCountryInfo {
pub fn new() -> Self {
Self {
info_by_country: BTreeMap::<String, BTreeMap<BridgeInfoType, u32>>::new(),
info_by_day: BTreeMap::<u32, BTreeMap<BridgeInfoType, u32>>::new(),
blocked: false,
}
}
pub fn add_info(
&mut self,
info_type: BridgeInfoType,
count_per_country: &BTreeMap<String, u32>,
) {
for country in count_per_country.keys() {
if self.info_by_country.contains_key(country) {
let info = self.info_by_country.get_mut(country).unwrap();
if !info.contains_key(&info_type) {
info.insert(
info_type,
*count_per_country.get(&country.to_string()).unwrap(),
);
} else if info_type == BridgeInfoType::BridgeIps {
// Use newest value we've seen today
if info.get(&info_type).unwrap() < count_per_country.get(country).unwrap() {
info.insert(
BridgeInfoType::BridgeIps,
*count_per_country.get(&country.to_string()).unwrap(),
);
}
} else {
let new_count = info.get(&info_type).unwrap()
+ *count_per_country.get(&country.to_string()).unwrap();
info.insert(info_type, new_count);
pub fn add_info(&mut self, info_type: BridgeInfoType, date: u32, count: u32) {
if self.info_by_day.contains_key(&date) {
let info = self.info_by_day.get_mut(&date).unwrap();
if !info.contains_key(&info_type) {
info.insert(info_type, count);
} else if info_type == BridgeInfoType::BridgeIps {
if *info.get(&info_type).unwrap() < count {
// Use highest value we've seen today
info.insert(info_type, count);
}
} else {
let mut info = BTreeMap::<BridgeInfoType, u32>::new();
info.insert(
info_type,
*count_per_country.get(&country.to_string()).unwrap(),
);
self.info_by_country.insert(country.to_string(), info);
// Add count to previous count for reports
let new_count = info.get(&info_type).unwrap() + count;
info.insert(info_type, new_count);
}
} else {
let mut info = BTreeMap::<BridgeInfoType, u32>::new();
info.insert(info_type, count);
self.info_by_day.insert(date, info);
}
}
}
impl fmt::Display for DailyBridgeInfo {
impl fmt::Display for BridgeCountryInfo {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
let mut str = String::from("info:");
for country in self.info_by_country.keys() {
let info = self.info_by_country.get(country).unwrap();
for date in self.info_by_day.keys() {
let info = self.info_by_day.get(date).unwrap();
let ip_count = match info.get(&BridgeInfoType::BridgeIps) {
Some(v) => v,
None => &0,
@ -170,8 +149,8 @@ impl fmt::Display for DailyBridgeInfo {
if ip_count > &0 || nr_count > &0 || pr_count > &0 {
str.push_str(
format!(
"\n cc: {}\n connections: {}\n negative reports: {}\n positive reports: {}",
country,
"\n date: {}\n connections: {}\n negative reports: {}\n positive reports: {}",
date,
ip_count,
nr_count,
pr_count,
@ -196,20 +175,29 @@ pub fn add_extra_info_to_db(db: &Db, extra_info: ExtraInfo) {
Some(v) => bincode::deserialize(&v).unwrap(),
None => BridgeInfo::new(fingerprint, &extra_info.nickname),
};
// If we already have an entry, compare it with the new one. For each
// country:count mapping, use the greater of the two counts.
if bridge_info.info_by_day.contains_key(&extra_info.date) {
let daily_bridge_info = bridge_info.info_by_day.get_mut(&extra_info.date).unwrap();
daily_bridge_info.add_info(BridgeInfoType::BridgeIps, &extra_info.bridge_ips);
} else {
// No existing entry; make a new one.
let mut daily_bridge_info = DailyBridgeInfo {
info_by_country: BTreeMap::<String, BTreeMap<BridgeInfoType, u32>>::new(),
};
daily_bridge_info.add_info(BridgeInfoType::BridgeIps, &extra_info.bridge_ips);
bridge_info
.info_by_day
.insert(extra_info.date, daily_bridge_info);
for country in extra_info.bridge_ips.keys() {
if bridge_info.info_by_country.contains_key::<String>(country) {
bridge_info
.info_by_country
.get_mut(country)
.unwrap()
.add_info(
BridgeInfoType::BridgeIps,
extra_info.date,
*extra_info.bridge_ips.get(country).unwrap(),
);
} else {
// No existing entry; make a new one.
let mut bridge_country_info = BridgeCountryInfo::new();
bridge_country_info.add_info(
BridgeInfoType::BridgeIps,
extra_info.date,
*extra_info.bridge_ips.get(country).unwrap(),
);
bridge_info
.info_by_country
.insert(country.to_string(), bridge_country_info);
}
}
// Commit changes to database
db.insert(fingerprint, bincode::serialize(&bridge_info).unwrap())
@ -260,11 +248,12 @@ pub fn save_negative_report_to_process(db: &Db, nr: NegativeReport) {
Some(v) => bincode::deserialize(&v).unwrap(),
None => BTreeMap::<String, BTreeMap<String, u32>>::new(),
};
// Store to-be-processed reports with key [fingerprint]_[date]
// Store to-be-processed reports with key [fingerprint]_[country]_[date]
let map_key = format!(
"{}_{}",
"{}_{}_{}",
array_bytes::bytes2hex("", &nr.fingerprint),
&nr.date
&nr.country,
&nr.date,
);
let serialized_nr = nr.to_json();
if reports.contains_key(&map_key) {
@ -322,12 +311,9 @@ pub async fn update_negative_reports(db: &Db, distributors: &BTreeMap<BridgeDist
Some(v) => bincode::deserialize(&v).unwrap(),
None => BTreeMap::<String, BTreeMap<String, u32>>::new(),
};
for bridge_date in all_negative_reports.keys() {
// We could parse the fingerprint and date:
//let fingerprint: [u8; 20] = array_bytes::hex2array(&bridge_date[0..40]).unwrap();
//let date: u32 = &bridge_date[41..].parse().unwrap();
// but instead, let's just get it from the first report
let reports = all_negative_reports.get(bridge_date).unwrap();
// Key is [fingerprint]_[country]_[date]
for bridge_country_date in all_negative_reports.keys() {
let reports = all_negative_reports.get(bridge_country_date).unwrap();
if !reports.is_empty() {
let first_report: SerializableNegativeReport =
serde_json::from_str(reports.first_key_value().unwrap().0).unwrap();
@ -335,8 +321,7 @@ pub async fn update_negative_reports(db: &Db, distributors: &BTreeMap<BridgeDist
let date = first_report.date;
let country = first_report.country;
let count_valid = verify_negative_reports(&distributors, reports).await;
let mut count_per_country = BTreeMap::<String, u32>::new();
count_per_country.insert(country, count_valid).unwrap();
let mut bridge_info = match db.get(&fingerprint).unwrap() {
Some(v) => bincode::deserialize(&v).unwrap(),
// It should already exist, unless the bridge hasn't published
@ -344,17 +329,19 @@ pub async fn update_negative_reports(db: &Db, distributors: &BTreeMap<BridgeDist
None => BridgeInfo::new(fingerprint, &"".to_string()),
};
// Add the new report count to it
if bridge_info.info_by_day.contains_key(&date) {
let daily_bridge_info = bridge_info.info_by_day.get_mut(&date).unwrap();
daily_bridge_info.add_info(BridgeInfoType::NegativeReports, &count_per_country);
if bridge_info.info_by_country.contains_key(&country) {
let bridge_country_info = bridge_info.info_by_country.get_mut(&country).unwrap();
bridge_country_info.add_info(BridgeInfoType::NegativeReports, date, count_valid);
// Commit changes to database
db.insert(fingerprint, bincode::serialize(&bridge_info).unwrap())
.unwrap();
} else {
// No existing entry; make a new one.
let mut daily_bridge_info = DailyBridgeInfo::new();
daily_bridge_info.add_info(BridgeInfoType::NegativeReports, &count_per_country);
bridge_info.info_by_day.insert(date, daily_bridge_info);
let mut bridge_country_info = BridgeCountryInfo::new();
bridge_country_info.add_info(BridgeInfoType::NegativeReports, date, count_valid);
bridge_info
.info_by_country
.insert(country, bridge_country_info);
// Commit changes to database
db.insert(fingerprint, bincode::serialize(&bridge_info).unwrap())
.unwrap();
@ -380,11 +367,12 @@ pub fn save_positive_report_to_process(db: &Db, pr: PositiveReport) {
Some(v) => bincode::deserialize(&v).unwrap(),
None => BTreeMap::<String, Vec<SerializablePositiveReport>>::new(),
};
// Store to-be-processed reports with key [fingerprint]_[date]
// Store to-be-processed reports with key [fingerprint]_[country]_[date]
let map_key = format!(
"{}_{}",
"{}_{}_{}",
array_bytes::bytes2hex("", &pr.fingerprint),
&pr.date
&pr.country,
&pr.date,
);
if reports.contains_key(&map_key) {
reports
@ -437,20 +425,15 @@ pub async fn update_positive_reports(db: &Db, distributors: &BTreeMap<BridgeDist
Some(v) => bincode::deserialize(&v).unwrap(),
None => BTreeMap::<String, Vec<SerializablePositiveReport>>::new(),
};
for bridge_date in all_positive_reports.keys() {
// We could parse the fingerprint and date:
//let fingerprint: [u8; 20] = array_bytes::hex2array(&bridge_date[0..40]).unwrap();
//let date: u32 = &bridge_date[41..].parse().unwrap();
// but instead, let's just get it from the first report
let reports = all_positive_reports.get(bridge_date).unwrap();
// Key is [fingerprint]_[country]_[date]
for bridge_country_date in all_positive_reports.keys() {
let reports = all_positive_reports.get(bridge_country_date).unwrap();
if !reports.is_empty() {
let first_report = &reports[0];
let fingerprint = first_report.fingerprint;
let date = first_report.date;
let country = first_report.country.clone();
let count_valid = verify_positive_reports(&distributors, reports).await;
let mut count_per_country = BTreeMap::<String, u32>::new();
count_per_country.insert(country, count_valid).unwrap();
let mut bridge_info = match db.get(&fingerprint).unwrap() {
Some(v) => bincode::deserialize(&v).unwrap(),
// It should already exist, unless the bridge hasn't published
@ -458,17 +441,19 @@ pub async fn update_positive_reports(db: &Db, distributors: &BTreeMap<BridgeDist
None => BridgeInfo::new(fingerprint, &"".to_string()),
};
// Add the new report count to it
if bridge_info.info_by_day.contains_key(&date) {
let daily_bridge_info = bridge_info.info_by_day.get_mut(&date).unwrap();
daily_bridge_info.add_info(BridgeInfoType::PositiveReports, &count_per_country);
if bridge_info.info_by_country.contains_key(&country) {
let bridge_country_info = bridge_info.info_by_country.get_mut(&country).unwrap();
bridge_country_info.add_info(BridgeInfoType::PositiveReports, date, count_valid);
// Commit changes to database
db.insert(fingerprint, bincode::serialize(&bridge_info).unwrap())
.unwrap();
} else {
// No existing entry; make a new one.
let mut daily_bridge_info = DailyBridgeInfo::new();
daily_bridge_info.add_info(BridgeInfoType::PositiveReports, &count_per_country);
bridge_info.info_by_day.insert(date, daily_bridge_info);
let mut bridge_country_info = BridgeCountryInfo::new();
bridge_country_info.add_info(BridgeInfoType::PositiveReports, date, count_valid);
bridge_info
.info_by_country
.insert(country, bridge_country_info);
// Commit changes to database
db.insert(fingerprint, bincode::serialize(&bridge_info).unwrap())
.unwrap();